Merge branch 'develop' into improve-freqai-docs

.github/workflows/ci.yml

@@ -272,6 +272,16 @@ jobs:
         pip install pyaml
         python build_helpers/pre_commit_update.py
 
+  pre-commit:
+    runs-on: ubuntu-22.04
+    steps:
+    - uses: actions/checkout@v3
+
+    - uses: actions/setup-python@v4
+      with:
+        python-version: "3.10"
+    - uses: pre-commit/action@v3.0.0
+
   docs_check:
     runs-on: ubuntu-20.04
     steps:
@@ -302,7 +312,7 @@ jobs:
 
   # Notify only once - when CI completes (and after deploy) in case it's successfull
   notify-complete:
-    needs: [ build_linux, build_macos, build_windows, docs_check, mypy_version_check ]
+    needs: [ build_linux, build_macos, build_windows, docs_check, mypy_version_check, pre-commit ]
     runs-on: ubuntu-20.04
     # Discord notification can't handle schedule events
     if: (github.event_name != 'schedule')
@@ -327,7 +337,7 @@ jobs:
           webhookUrl: ${{ secrets.DISCORD_WEBHOOK }}
 
   deploy:
-    needs: [ build_linux, build_macos, build_windows, docs_check, mypy_version_check ]
+    needs: [ build_linux, build_macos, build_windows, docs_check, mypy_version_check, pre-commit ]
     runs-on: ubuntu-20.04
 
     if: (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'release') && github.repository == 'freqtrade/freqtrade'

config_examples/config_full.example.json

@@ -172,7 +172,24 @@
         "jwt_secret_key": "somethingrandom",
         "CORS_origins": [],
         "username": "freqtrader",
-        "password": "SuperSecurePassword"
+        "password": "SuperSecurePassword",
+        "ws_token": "secret_ws_t0ken."
+    },
+    "external_message_consumer": {
+        "enabled": false,
+        "producers": [
+          {
+            "name": "default",
+            "host": "127.0.0.2",
+            "port": 8080,
+            "ws_token": "secret_ws_t0ken."
+          }
+        ],
+        "wait_timeout": 300,
+        "ping_timeout": 10,
+        "sleep_time": 10,
+        "remove_entry_exit_signals": false,
+        "message_size_limit": 8
     },
     "bot_name": "freqtrade",
     "db_url": "sqlite:///tradesv3.sqlite",

docker/Dockerfile.freqai

@@ -6,4 +6,3 @@ FROM ${sourceimage}:${sourcetag}
 COPY requirements-freqai.txt /freqtrade/
 
 RUN pip install -r requirements-freqai.txt --user --no-cache-dir
-

docs/configuration.md

@@ -225,14 +225,16 @@ Mandatory parameters are marked as **Required**, which means that they are requi
 | `webhook.webhookexitcancel` | Payload to send on exit order cancel. Only required if `webhook.enabled` is `true`. See the [webhook documentation](webhook-config.md) for more details. <br> **Datatype:** String
 | `webhook.webhookexitfill` | Payload to send on exit order filled. Only required if `webhook.enabled` is `true`. See the [webhook documentation](webhook-config.md) for more details. <br> **Datatype:** String
 | `webhook.webhookstatus` | Payload to send on status calls. Only required if `webhook.enabled` is `true`. See the [webhook documentation](webhook-config.md) for more details. <br> **Datatype:** String
-| | **Rest API / FreqUI**
+| | **Rest API / FreqUI / Producer-Consumer**
 | `api_server.enabled` | Enable usage of API Server. See the [API Server documentation](rest-api.md) for more details. <br> **Datatype:** Boolean
 | `api_server.listen_ip_address` | Bind IP address. See the [API Server documentation](rest-api.md) for more details. <br> **Datatype:** IPv4
 | `api_server.listen_port` | Bind Port. See the [API Server documentation](rest-api.md) for more details. <br>**Datatype:** Integer between 1024 and 65535
 | `api_server.verbosity` | Logging verbosity. `info` will print all RPC Calls, while "error" will only display errors. <br>**Datatype:** Enum, either `info` or `error`. Defaults to `info`.
 | `api_server.username` | Username for API server. See the [API Server documentation](rest-api.md) for more details. <br>**Keep it in secret, do not disclose publicly.**<br> **Datatype:** String
 | `api_server.password` | Password for API server. See the [API Server documentation](rest-api.md) for more details. <br>**Keep it in secret, do not disclose publicly.**<br> **Datatype:** String
+| `api_server.ws_token` | API token for the Message WebSocket. See the [API Server documentation](rest-api.md) for more details.  <br>**Keep it in secret, do not disclose publicly.** <br> **Datatype:** String
 | `bot_name` | Name of the bot. Passed via API to a client - can be shown to distinguish / name bots.<br> *Defaults to `freqtrade`*<br> **Datatype:** String
+| `external_message_consumer` | Enable [Producer/Consumer mode](producer-consumer.md) for more details. <br> **Datatype:** Dict
 | | **Other**
 | `initial_state` | Defines the initial application state. If set to stopped, then the bot has to be explicitly started via `/start` RPC command. <br>*Defaults to `stopped`.* <br> **Datatype:** Enum, either `stopped` or `running`
 | `force_entry_enable` | Enables the RPC Commands to force a Trade entry. More information below. <br> **Datatype:** Boolean

docs/data-download.md

@@ -26,7 +26,7 @@ usage: freqtrade download-data [-h] [-v] [--logfile FILE] [-V] [-c PATH]
                                [--timerange TIMERANGE] [--dl-trades]
                                [--exchange EXCHANGE]
                                [-t TIMEFRAMES [TIMEFRAMES ...]] [--erase]
-                               [--data-format-ohlcv {json,jsongz,hdf5}]
+                               [--data-format-ohlcv {json,jsongz,hdf5,feather,parquet}]
                                [--data-format-trades {json,jsongz,hdf5}]
                                [--trading-mode {spot,margin,futures}]
                                [--prepend]
@@ -55,7 +55,7 @@ optional arguments:
                         list. Default: `1m 5m`.
   --erase               Clean all existing data for the selected
                         exchange/pairs/timeframes.
-  --data-format-ohlcv {json,jsongz,hdf5}
+  --data-format-ohlcv {json,jsongz,hdf5,feather,parquet}
                         Storage format for downloaded candle (OHLCV) data.
                         (default: `json`).
   --data-format-trades {json,jsongz,hdf5}
@@ -76,7 +76,7 @@ Common arguments:
                         `userdir/config.json` or `config.json` whichever
                         exists). Multiple --config options may be used. Can be
                         set to `-` to read config from stdin.
-  -d PATH, --datadir PATH
+  -d PATH, --datadir PATH, --data-dir PATH
                         Path to directory with historical backtesting data.
   --userdir PATH, --user-data-dir PATH
                         Path to userdata directory.
@@ -179,9 +179,11 @@ freqtrade download-data --exchange binance --pairs ETH/USDT XRP/USDT BTC/USDT --
 
 Freqtrade currently supports 3 data-formats for both OHLCV and trades data:
 
-* `json` (plain "text" json files)
+* `json` -  plain "text" json files
-* `jsongz` (a gzip-zipped version of json files)
+* `jsongz` - a gzip-zipped version of json files
-* `hdf5` (a high performance datastore)
+* `hdf5` - a high performance datastore
+* `feather` - a dataformat based on Apache Arrow
+* `parquet` - columnar datastore
 
 By default, OHLCV data is stored as `json` data, while trades data is stored as `jsongz` data.
 
@@ -200,38 +202,74 @@ If the default data-format has been changed during download, then the keys `data
 !!! Note
     You can convert between data-formats using the [convert-data](#sub-command-convert-data) and [convert-trade-data](#sub-command-convert-trade-data) methods.
 
+#### Dataformat comparison
+
+The following comparisons have been made with the following data, and by using the linux `time` command.
+
+```
+Found 6 pair / timeframe combinations.
++----------+-------------+--------+---------------------+---------------------+
+|     Pair |   Timeframe |   Type |                From |                  To |
+|----------+-------------+--------+---------------------+---------------------|
+| BTC/USDT |          5m |   spot | 2017-08-17 04:00:00 | 2022-09-13 19:25:00 |
+| ETH/USDT |          1m |   spot | 2017-08-17 04:00:00 | 2022-09-13 19:26:00 |
+| BTC/USDT |          1m |   spot | 2017-08-17 04:00:00 | 2022-09-13 19:30:00 |
+| XRP/USDT |          5m |   spot | 2018-05-04 08:10:00 | 2022-09-13 19:15:00 |
+| XRP/USDT |          1m |   spot | 2018-05-04 08:11:00 | 2022-09-13 19:22:00 |
+| ETH/USDT |          5m |   spot | 2017-08-17 04:00:00 | 2022-09-13 19:20:00 |
++----------+-------------+--------+---------------------+---------------------+
+```
+
+Timings have been taken in a not very scientific way with the following command, which forces reading the data into memory.
+
+``` bash
+time freqtrade list-data --show-timerange --data-format-ohlcv <dataformat>
+```
+
+|  Format | Size | timing |
+|------------|-------------|-------------|
+| `json` | 149Mb | 25.6s |
+| `jsongz` | 39Mb | 27s |
+| `hdf5` | 145Mb | 3.9s |
+| `feather` | 72Mb | 3.5s |
+| `parquet` | 83Mb | 3.8s |
+
+Size has been taken from the BTC/USDT 1m spot combination for the timerange specified above.
+
+To have a best performance/size mix, we recommend the use of either feather or parquet.
+
 #### Sub-command convert data
 
 ```
 usage: freqtrade convert-data [-h] [-v] [--logfile FILE] [-V] [-c PATH]
                               [-d PATH] [--userdir PATH]
                               [-p PAIRS [PAIRS ...]] --format-from
-                              {json,jsongz,hdf5} --format-to
+                              {json,jsongz,hdf5,feather,parquet} --format-to
-                              {json,jsongz,hdf5} [--erase]
+                              {json,jsongz,hdf5,feather,parquet} [--erase]
-                              [-t {1m,3m,5m,15m,30m,1h,2h,4h,6h,8h,12h,1d,3d,1w,2w,1M,1y} [{1m,3m,5m,15m,30m,1h,2h,4h,6h,8h,12h,1d,3d,1w,2w,1M,1y} ...]]
                               [--exchange EXCHANGE]
+                              [-t TIMEFRAMES [TIMEFRAMES ...]]
                               [--trading-mode {spot,margin,futures}]
-                              [--candle-types {spot,,futures,mark,index,premiumIndex,funding_rate} [{spot,,futures,mark,index,premiumIndex,funding_rate} ...]]
+                              [--candle-types {spot,futures,mark,index,premiumIndex,funding_rate} [{spot,futures,mark,index,premiumIndex,funding_rate} ...]]
 
 optional arguments:
   -h, --help            show this help message and exit
   -p PAIRS [PAIRS ...], --pairs PAIRS [PAIRS ...]
                         Limit command to these pairs. Pairs are space-
                         separated.
-  --format-from {json,jsongz,hdf5}
+  --format-from {json,jsongz,hdf5,feather,parquet}
                         Source format for data conversion.
-  --format-to {json,jsongz,hdf5}
+  --format-to {json,jsongz,hdf5,feather,parquet}
                         Destination format for data conversion.
   --erase               Clean all existing data for the selected
                         exchange/pairs/timeframes.
-  -t {1m,3m,5m,15m,30m,1h,2h,4h,6h,8h,12h,1d,3d,1w,2w,1M,1y} [{1m,3m,5m,15m,30m,1h,2h,4h,6h,8h,12h,1d,3d,1w,2w,1M,1y} ...], --timeframes {1m,3m,5m,15m,30m,1h,2h,4h,6h,8h,12h,1d,3d,1w,2w,1M,1y} [{1m,3m,5m,15m,30m,1h,2h,4h,6h,8h,12h,1d,3d,1w,2w,1M,1y} ...]
-                        Specify which tickers to download. Space-separated
-                        list. Default: `1m 5m`.
   --exchange EXCHANGE   Exchange name (default: `bittrex`). Only valid if no
                         config is provided.
-  --trading-mode {spot,margin,futures}
+  -t TIMEFRAMES [TIMEFRAMES ...], --timeframes TIMEFRAMES [TIMEFRAMES ...]
+                        Specify which tickers to download. Space-separated
+                        list. Default: `1m 5m`.
+  --trading-mode {spot,margin,futures}, --tradingmode {spot,margin,futures}
                         Select Trading mode
-  --candle-types {spot,,futures,mark,index,premiumIndex,funding_rate} [{spot,,futures,mark,index,premiumIndex,funding_rate} ...]
+  --candle-types {spot,futures,mark,index,premiumIndex,funding_rate} [{spot,futures,mark,index,premiumIndex,funding_rate} ...]
                         Select candle type to use
 
 Common arguments:
@@ -245,7 +283,7 @@ Common arguments:
                         `userdir/config.json` or `config.json` whichever
                         exists). Multiple --config options may be used. Can be
                         set to `-` to read config from stdin.
-  -d PATH, --datadir PATH
+  -d PATH, --datadir PATH, --data-dir PATH
                         Path to directory with historical backtesting data.
   --userdir PATH, --user-data-dir PATH
                         Path to userdata directory.
@@ -267,20 +305,24 @@ freqtrade convert-data --format-from json --format-to jsongz --datadir ~/.freqtr
 usage: freqtrade convert-trade-data [-h] [-v] [--logfile FILE] [-V] [-c PATH]
                                     [-d PATH] [--userdir PATH]
                                     [-p PAIRS [PAIRS ...]] --format-from
-                                    {json,jsongz,hdf5} --format-to
+                                    {json,jsongz,hdf5,feather,parquet}
-                                    {json,jsongz,hdf5} [--erase]
+                                    --format-to
+                                    {json,jsongz,hdf5,feather,parquet}
+                                    [--erase] [--exchange EXCHANGE]
 
 optional arguments:
   -h, --help            show this help message and exit
   -p PAIRS [PAIRS ...], --pairs PAIRS [PAIRS ...]
-                        Show profits for only these pairs. Pairs are space-
+                        Limit command to these pairs. Pairs are space-
                         separated.
-  --format-from {json,jsongz,hdf5}
+  --format-from {json,jsongz,hdf5,feather,parquet}
                         Source format for data conversion.
-  --format-to {json,jsongz,hdf5}
+  --format-to {json,jsongz,hdf5,feather,parquet}
                         Destination format for data conversion.
   --erase               Clean all existing data for the selected
                         exchange/pairs/timeframes.
+  --exchange EXCHANGE   Exchange name (default: `bittrex`). Only valid if no
+                        config is provided.
 
 Common arguments:
   -v, --verbose         Verbose mode (-vv for more, -vvv to get all messages).
@@ -293,7 +335,7 @@ Common arguments:
                         `userdir/config.json` or `config.json` whichever
                         exists). Multiple --config options may be used. Can be
                         set to `-` to read config from stdin.
-  -d PATH, --datadir PATH
+  -d PATH, --datadir PATH, --data-dir PATH
                         Path to directory with historical backtesting data.
   --userdir PATH, --user-data-dir PATH
                         Path to userdata directory.
@@ -318,9 +360,9 @@ This command will allow you to repeat this last step for additional timeframes w
 usage: freqtrade trades-to-ohlcv [-h] [-v] [--logfile FILE] [-V] [-c PATH]
                                  [-d PATH] [--userdir PATH]
                                  [-p PAIRS [PAIRS ...]]
-                                 [-t {1m,3m,5m,15m,30m,1h,2h,4h,6h,8h,12h,1d,3d,1w,2w,1M,1y} [{1m,3m,5m,15m,30m,1h,2h,4h,6h,8h,12h,1d,3d,1w,2w,1M,1y} ...]]
+                                 [-t TIMEFRAMES [TIMEFRAMES ...]]
                                  [--exchange EXCHANGE]
-                                 [--data-format-ohlcv {json,jsongz,hdf5}]
+                                 [--data-format-ohlcv {json,jsongz,hdf5,feather,parquet}]
                                  [--data-format-trades {json,jsongz,hdf5}]
 
 optional arguments:
@@ -328,12 +370,12 @@ optional arguments:
   -p PAIRS [PAIRS ...], --pairs PAIRS [PAIRS ...]
                         Limit command to these pairs. Pairs are space-
                         separated.
-  -t {1m,3m,5m,15m,30m,1h,2h,4h,6h,8h,12h,1d,3d,1w,2w,1M,1y} [{1m,3m,5m,15m,30m,1h,2h,4h,6h,8h,12h,1d,3d,1w,2w,1M,1y} ...], --timeframes {1m,3m,5m,15m,30m,1h,2h,4h,6h,8h,12h,1d,3d,1w,2w,1M,1y} [{1m,3m,5m,15m,30m,1h,2h,4h,6h,8h,12h,1d,3d,1w,2w,1M,1y} ...]
+  -t TIMEFRAMES [TIMEFRAMES ...], --timeframes TIMEFRAMES [TIMEFRAMES ...]
                         Specify which tickers to download. Space-separated
                         list. Default: `1m 5m`.
   --exchange EXCHANGE   Exchange name (default: `bittrex`). Only valid if no
                         config is provided.
-  --data-format-ohlcv {json,jsongz,hdf5}
+  --data-format-ohlcv {json,jsongz,hdf5,feather,parquet}
                         Storage format for downloaded candle (OHLCV) data.
                         (default: `json`).
   --data-format-trades {json,jsongz,hdf5}
@@ -351,7 +393,7 @@ Common arguments:
                         `userdir/config.json` or `config.json` whichever
                         exists). Multiple --config options may be used. Can be
                         set to `-` to read config from stdin.
-  -d PATH, --datadir PATH
+  -d PATH, --datadir PATH, --data-dir PATH
                         Path to directory with historical backtesting data.
   --userdir PATH, --user-data-dir PATH
                         Path to userdata directory.
@@ -371,7 +413,7 @@ You can get a list of downloaded data using the `list-data` sub-command.
 ```
 usage: freqtrade list-data [-h] [-v] [--logfile FILE] [-V] [-c PATH] [-d PATH]
                            [--userdir PATH] [--exchange EXCHANGE]
-                           [--data-format-ohlcv {json,jsongz,hdf5}]
+                           [--data-format-ohlcv {json,jsongz,hdf5,feather,parquet}]
                            [-p PAIRS [PAIRS ...]]
                            [--trading-mode {spot,margin,futures}]
                            [--show-timerange]
@@ -380,13 +422,13 @@ optional arguments:
   -h, --help            show this help message and exit
   --exchange EXCHANGE   Exchange name (default: `bittrex`). Only valid if no
                         config is provided.
-  --data-format-ohlcv {json,jsongz,hdf5}
+  --data-format-ohlcv {json,jsongz,hdf5,feather,parquet}
                         Storage format for downloaded candle (OHLCV) data.
                         (default: `json`).
   -p PAIRS [PAIRS ...], --pairs PAIRS [PAIRS ...]
                         Limit command to these pairs. Pairs are space-
                         separated.
-  --trading-mode {spot,margin,futures}
+  --trading-mode {spot,margin,futures}, --tradingmode {spot,margin,futures}
                         Select Trading mode
   --show-timerange      Show timerange available for available data. (May take
                         a while to calculate).
@@ -402,7 +444,7 @@ Common arguments:
                         `userdir/config.json` or `config.json` whichever
                         exists). Multiple --config options may be used. Can be
                         set to `-` to read config from stdin.
-  -d PATH, --datadir PATH
+  -d PATH, --datadir PATH, --data-dir PATH
                         Path to directory with historical backtesting data.
   --userdir PATH, --user-data-dir PATH
                         Path to userdata directory.

docs/producer-consumer.md

@@ -0,0 +1,163 @@
+# Producer / Consumer mode
+
+freqtrade provides a mechanism whereby an instance (also called `consumer`) may listen to messages from an upstream freqtrade instance (also called `producer`) using the message websocket. Mainly, `analyzed_df` and `whitelist` messages. This allows the reuse of computed indicators (and signals) for pairs in multiple bots without needing to compute them multiple times.
+
+See [Message Websocket](rest-api.md#message-websocket) in the Rest API docs for setting up the `api_server` configuration for your message websocket (this will be your producer).
+
+!!! Note
+    We strongly recommend to set `ws_token` to something random and known only to yourself to avoid unauthorized access to your bot.
+
+## Configuration
+
+Enable subscribing to an instance by adding the `external_message_consumer` section to the consumer's config file.
+
+```json
+{
+    //...
+   "external_message_consumer": {
+        "enabled": true,
+        "producers": [
+            {
+                "name": "default", // This can be any name you'd like, default is "default"
+                "host": "127.0.0.1", // The host from your producer's api_server config
+                "port": 8080, // The port from your producer's api_server config
+                "ws_token": "sercet_Ws_t0ken" // The ws_token from your producer's api_server config
+            }
+        ],
+        // The following configurations are optional, and usually not required
+        // "wait_timeout": 300,
+        // "ping_timeout": 10,
+        // "sleep_time": 10,
+        // "remove_entry_exit_signals": false,
+        // "message_size_limit": 8
+    }
+    //...
+}
+```
+
+|  Parameter | Description |
+|------------|-------------|
+| `enabled` | **Required.** Enable consumer mode. If set to false, all other settings in this section are ignored.<br>*Defaults to `false`.*<br> **Datatype:** boolean .
+| `producers` | **Required.** List of producers <br> **Datatype:** Array.
+| `producers.name` | **Required.** Name of this producer. This name must be used in calls to `get_producer_pairs()` and `get_producer_df()` if more than one producer is used.<br> **Datatype:** string
+| `producers.host` | **Required.** The hostname or IP address from your producer.<br> **Datatype:** string
+| `producers.port` | **Required.** The port matching the above host.<br> **Datatype:** string
+| `producers.ws_token` | **Required.**  `ws_token` as configured on the producer.<br> **Datatype:** string
+| | **Optional settings**
+| `wait_timeout` | Timeout until we ping again if no message is received. <br>*Defaults to `300`.*<br> **Datatype:** Integer - in seconds.
+| `wait_timeout` | Ping timeout <br>*Defaults to `10`.*<br> **Datatype:** Integer - in seconds.
+| `sleep_time` | Sleep time before retrying to connect.<br>*Defaults to `10`.*<br> **Datatype:** Integer - in seconds.
+| `remove_entry_exit_signals` | Remove signal columns from the dataframe (set them to 0) on dataframe receipt.<br>*Defaults to `10`.*<br> **Datatype:** Integer - in seconds.
+| `message_size_limit` | Size limit per message<br>*Defaults to `8`.*<br> **Datatype:** Integer - Megabytes.
+
+Instead of (or as well as) calculating indicators in `populate_indicators()` the follower instance listens on the connection to a producer instance's messages (or multiple producer instances in advanced configurations) and requests the producer's most recently analyzed dataframes for each pair in the active whitelist.
+
+A consumer instance will then have a full copy of the analyzed dataframes without the need to calculate them itself.
+
+## Examples
+
+### Example - Producer Strategy
+
+A simple strategy with multiple indicators. No special considerations are required in the strategy itself.
+
+```py
+class ProducerStrategy(IStrategy):
+    #...
+    def populate_indicators(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        """
+        Calculate indicators in the standard freqtrade way which can then be broadcast to other instances
+        """
+        dataframe['rsi'] = ta.RSI(dataframe)
+        bollinger = qtpylib.bollinger_bands(qtpylib.typical_price(dataframe), window=20, stds=2)
+        dataframe['bb_lowerband'] = bollinger['lower']
+        dataframe['bb_middleband'] = bollinger['mid']
+        dataframe['bb_upperband'] = bollinger['upper']
+        dataframe['tema'] = ta.TEMA(dataframe, timeperiod=9)
+
+        return dataframe
+
+    def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        """
+        Populates the entry signal for the given dataframe
+        """
+        dataframe.loc[
+            (
+                (qtpylib.crossed_above(dataframe['rsi'], self.buy_rsi.value)) &
+                (dataframe['tema'] <= dataframe['bb_middleband']) &
+                (dataframe['tema'] > dataframe['tema'].shift(1)) &
+                (dataframe['volume'] > 0)
+            ),
+            'enter_long'] = 1
+
+        return dataframe
+```
+
+!!! Tip "FreqAI"
+    You can use this to setup [FreqAI](freqai.md) on a powerful machine, while you run consumers on simple machines like raspberries, which can interpret the signals generated from the producer in different ways.
+
+
+### Example - Consumer Strategy
+
+A logically equivalent strategy which calculates no indicators itself, but will have the same analyzed dataframes available to make trading decisions based on the indicators calculated in the producer. In this example the consumer has the same entry criteria, however this is not necessary. The consumer may use different logic to enter/exit trades, and only use the indicators as specified.
+
+```py
+class ConsumerStrategy(IStrategy):
+    #...
+    process_only_new_candles = False # required for consumers
+
+    _columns_to_expect = ['rsi_default', 'tema_default', 'bb_middleband_default']
+
+    def populate_indicators(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        """
+        Use the websocket api to get pre-populated indicators from another freqtrade instance.
+        Use `self.dp.get_producer_df(pair)` to get the dataframe
+        """
+        pair = metadata['pair']
+        timeframe = self.timeframe
+
+        producer_pairs = self.dp.get_producer_pairs()
+        # You can specify which producer to get pairs from via:
+        # self.dp.get_producer_pairs("my_other_producer")
+
+        # This func returns the analyzed dataframe, and when it was analyzed
+        producer_dataframe, _ = self.dp.get_producer_df(pair)
+        # You can get other data if the producer makes it available:
+        # self.dp.get_producer_df(
+        #   pair,
+        #   timeframe="1h",
+        #   candle_type=CandleType.SPOT,
+        #   producer_name="my_other_producer"
+        # )
+
+        if not producer_dataframe.empty:
+            # If you plan on passing the producer's entry/exit signal directly,
+            # specify ffill=False or it will have unintended results
+            merged_dataframe = merge_informative_pair(dataframe, producer_dataframe,
+                                                      timeframe, timeframe,
+                                                      append_timeframe=False,
+                                                      suffix="default")
+            return merged_dataframe
+        else:
+            dataframe[self._columns_to_expect] = 0
+
+        return dataframe
+
+    def populate_entry_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
+        """
+        Populates the entry signal for the given dataframe
+        """
+        # Use the dataframe columns as if we calculated them ourselves
+        dataframe.loc[
+            (
+                (qtpylib.crossed_above(dataframe['rsi_default'], self.buy_rsi.value)) &
+                (dataframe['tema_default'] <= dataframe['bb_middleband_default']) &
+                (dataframe['tema_default'] > dataframe['tema_default'].shift(1)) &
+                (dataframe['volume'] > 0)
+            ),
+            'enter_long'] = 1
+
+        return dataframe
+```
+
+!!! Tip "Using upstream signals"
+    By setting `remove_entry_exit_signals=false`, you can also use the producer's signals directly. They should be available as `enter_long_default` (assuming `suffix="default"` was used) - and can be used as either signal directly, or as additional indicator.

docs/rest-api.md

@@ -31,7 +31,8 @@ Sample configuration:
         "jwt_secret_key": "somethingrandom",
         "CORS_origins": [],
         "username": "Freqtrader",
-        "password": "SuperSecret1!"
+        "password": "SuperSecret1!",
+        "ws_token": "sercet_Ws_t0ken"
     },
 ```
 
@@ -93,7 +94,6 @@ Make sure that the following 2 lines are available in your docker-compose file:
 !!! Danger "Security warning"
     By using `8080:8080` in the docker port mapping, the API will be available to everyone connecting to the server under the correct port, so others may be able to control your bot.
 
-
 ## Rest API
 
 ### Consuming the API
@@ -322,6 +322,73 @@ whitelist
 
 ```
 
+### Message WebSocket
+
+The API Server includes a websocket endpoint for subscribing to RPC messages from the freqtrade Bot.
+This can be used to consume real-time data from your bot, such as entry/exit fill messages, whitelist changes, populated indicators for pairs, and more.
+
+This is also used to setup [Producer/Consumer mode](producer-consumer.md) in Freqtrade.
+
+Assuming your rest API is set to `127.0.0.1` on port `8080`, the endpoint is available at `http://localhost:8080/api/v1/message/ws`.
+
+To access the websocket endpoint, the `ws_token` is required as a query parameter in the endpoint URL.
+
+To generate a safe `ws_token` you can run the following code:
+
+``` python
+>>> import secrets
+>>> secrets.token_urlsafe(25)
+'hZ-y58LXyX_HZ8O1cJzVyN6ePWrLpNQv4Q'
+```
+
+You would then add that token under `ws_token` in your `api_server` config. Like so:
+
+``` json
+"api_server": {
+    "enabled": true,
+    "listen_ip_address": "127.0.0.1",
+    "listen_port": 8080,
+    "verbosity": "error",
+    "enable_openapi": false,
+    "jwt_secret_key": "somethingrandom",
+    "CORS_origins": [],
+    "username": "Freqtrader",
+    "password": "SuperSecret1!",
+    "ws_token": "hZ-y58LXyX_HZ8O1cJzVyN6ePWrLpNQv4Q" // <-----
+},
+```
+
+You can now connect to the endpoint at `http://localhost:8080/api/v1/message/ws?token=hZ-y58LXyX_HZ8O1cJzVyN6ePWrLpNQv4Q`.
+
+!!! Danger "Reuse of example tokens"
+    Please do not use the above example token. To make sure you are secure, generate a completely new token.
+
+#### Using the WebSocket
+
+Once connected to the WebSocket, the bot will broadcast RPC messages to anyone who is subscribed to them. To subscribe to a list of messages, you must send a JSON request through the WebSocket like the one below. The `data` key must be a list of message type strings.
+
+``` json
+{
+  "type": "subscribe",
+  "data": ["whitelist", "analyzed_df"] // A list of string message types
+}
+```
+
+For a list of message types, please refer to the RPCMessageType enum in `freqtrade/enums/rpcmessagetype.py`
+
+Now anytime those types of RPC messages are sent in the bot, you will receive them through the WebSocket as long as the connection is active. They typically take the same form as the request:
+
+``` json
+{
+  "type": "analyzed_df",
+  "data": {
+      "key": ["NEO/BTC", "5m", "spot"],
+      "df": {}, // The dataframe
+      "la": "2022-09-08 22:14:41.457786+00:00"
+  }
+}
+```
+
 ### OpenAPI interface
 
 To enable the builtin openAPI interface (Swagger UI), specify `"enable_openapi": true` in the api_server configuration.

docs/strategy-advanced.md

@@ -106,6 +106,12 @@ def custom_exit(self, pair: str, trade: Trade, current_time: datetime, current_r
 !!! Note
     `enter_tag` is limited to 100 characters, remaining data will be truncated.
 
+!!! Warning
+    There is only one `enter_tag` column, which is used for both long and short trades.
+    As a consequence, this column must be treated as "last write wins" (it's just a dataframe column after all).
+    In fancy situations, where multiple signals collide (or if signals are deactivated again based on different conditions), this can lead to odd results with the wrong tag applied to an entry signal.
+    These results are a consequence of the strategy overwriting prior tags - where the last tag will "stick" and will be the one freqtrade will use.
+
 ## Exit tag
 
 Similar to [Buy Tagging](#buy-tag), you can also specify a sell tag.

environment.yml

@@ -34,6 +34,7 @@ dependencies:
     - schedule
     - python-dateutil
     - joblib
+    - pyarrow
 
 
     # ============================

freqtrade/commands/build_config_commands.py

@@ -211,6 +211,7 @@ def ask_user_config() -> Dict[str, Any]:
     )
     # Force JWT token to be a random string
     answers['api_server_jwt_key'] = secrets.token_hex()
+    answers['api_server_ws_token'] = secrets.token_urlsafe(25)
 
     return answers
 

freqtrade/commands/cli_options.py

@@ -440,7 +440,7 @@ AVAILABLE_CLI_OPTIONS = {
     "dataformat_trades": Arg(
         '--data-format-trades',
         help='Storage format for downloaded trades data. (default: `jsongz`).',
-        choices=constants.AVAILABLE_DATAHANDLERS,
+        choices=constants.AVAILABLE_DATAHANDLERS_TRADES,
     ),
     "show_timerange": Arg(
         '--show-timerange',

freqtrade/configuration/config_validation.py

@@ -1,4 +1,5 @@
 import logging
+from collections import Counter
 from copy import deepcopy
 from typing import Any, Dict
 
@@ -85,6 +86,7 @@ def validate_config_consistency(conf: Dict[str, Any], preliminary: bool = False)
     _validate_unlimited_amount(conf)
     _validate_ask_orderbook(conf)
     _validate_freqai_hyperopt(conf)
+    _validate_consumers(conf)
     validate_migrated_strategy_settings(conf)
 
     # validate configuration before returning
@@ -332,6 +334,23 @@ def _validate_freqai_hyperopt(conf: Dict[str, Any]) -> None:
             'Using analyze-per-epoch parameter is not supported with a FreqAI strategy.')
 
 
+def _validate_consumers(conf: Dict[str, Any]) -> None:
+    emc_conf = conf.get('external_message_consumer', {})
+    if emc_conf.get('enabled', False):
+        if len(emc_conf.get('producers', [])) < 1:
+            raise OperationalException("You must specify at least 1 Producer to connect to.")
+
+        producer_names = [p['name'] for p in emc_conf.get('producers', [])]
+        duplicates = [item for item, count in Counter(producer_names).items() if count > 1]
+        if duplicates:
+            raise OperationalException(
+                f"Producer names must be unique. Duplicate: {', '.join(duplicates)}")
+        if conf.get('process_only_new_candles', True):
+            # Warning here or require it?
+            logger.warning("To receive best performance with external data, "
+                           "please set `process_only_new_candles` to False")
+
+
 def _strategy_settings(conf: Dict[str, Any]) -> None:
 
     process_deprecated_setting(conf, None, 'use_sell_signal', None, 'use_exit_signal')

freqtrade/constants.py

@@ -36,7 +36,8 @@ AVAILABLE_PAIRLISTS = ['StaticPairList', 'VolumePairList',
                        'PrecisionFilter', 'PriceFilter', 'RangeStabilityFilter',
                        'ShuffleFilter', 'SpreadFilter', 'VolatilityFilter']
 AVAILABLE_PROTECTIONS = ['CooldownPeriod', 'LowProfitPairs', 'MaxDrawdown', 'StoplossGuard']
-AVAILABLE_DATAHANDLERS = ['json', 'jsongz', 'hdf5']
+AVAILABLE_DATAHANDLERS_TRADES = ['json', 'jsongz', 'hdf5']
+AVAILABLE_DATAHANDLERS = AVAILABLE_DATAHANDLERS_TRADES + ['feather', 'parquet']
 BACKTEST_BREAKDOWNS = ['day', 'week', 'month']
 BACKTEST_CACHE_AGE = ['none', 'day', 'week', 'month']
 BACKTEST_CACHE_DEFAULT = 'day'
@@ -243,6 +244,7 @@ CONF_SCHEMA = {
         'exchange': {'$ref': '#/definitions/exchange'},
         'edge': {'$ref': '#/definitions/edge'},
         'freqai': {'$ref': '#/definitions/freqai'},
+        'external_message_consumer': {'$ref': '#/definitions/external_message_consumer'},
         'experimental': {
             'type': 'object',
             'properties': {
@@ -404,6 +406,7 @@ CONF_SCHEMA = {
                 },
                 'username': {'type': 'string'},
                 'password': {'type': 'string'},
+                'ws_token': {'type': ['string', 'array'], 'items': {'type': 'string'}},
                 'jwt_secret_key': {'type': 'string'},
                 'CORS_origins': {'type': 'array', 'items': {'type': 'string'}},
                 'verbosity': {'type': 'string', 'enum': ['error', 'info']},
@@ -432,7 +435,7 @@ CONF_SCHEMA = {
         },
         'dataformat_trades': {
             'type': 'string',
-            'enum': AVAILABLE_DATAHANDLERS,
+            'enum': AVAILABLE_DATAHANDLERS_TRADES,
             'default': 'jsongz'
         },
         'position_adjustment_enable': {'type': 'boolean'},
@@ -488,6 +491,47 @@ CONF_SCHEMA = {
             },
             'required': ['process_throttle_secs', 'allowed_risk']
         },
+        'external_message_consumer': {
+            'type': 'object',
+            'properties': {
+                'enabled': {'type': 'boolean', 'default': False},
+                'producers': {
+                    'type': 'array',
+                    'items': {
+                        'type': 'object',
+                        'properties': {
+                            'name': {'type': 'string'},
+                            'host': {'type': 'string'},
+                            'port': {
+                                'type': 'integer',
+                                'default': 8080,
+                                'minimum': 0,
+                                'maximum': 65535
+                            },
+                            'ws_token': {'type': 'string'},
+                        },
+                        'required': ['name', 'host', 'ws_token']
+                    }
+                },
+                'wait_timeout': {'type': 'integer', 'minimum': 0},
+                'sleep_time': {'type': 'integer', 'minimum': 0},
+                'ping_timeout': {'type': 'integer', 'minimum': 0},
+                'remove_entry_exit_signals': {'type': 'boolean', 'default': False},
+                'initial_candle_limit': {
+                    'type': 'integer',
+                    'minimum': 0,
+                    'maximum': 1500,
+                    'default': 1500
+                },
+                'message_size_limit': {  # In megabytes
+                    'type': 'integer',
+                    'minimum': 1,
+                    'maxmium': 20,
+                    'default': 8,
+                }
+            },
+            'required': ['producers']
+        },
         "freqai": {
             "type": "object",
             "properties": {

freqtrade/data/dataprovider.py

@@ -14,9 +14,10 @@ from pandas import DataFrame
 from freqtrade.configuration import TimeRange
 from freqtrade.constants import Config, ListPairsWithTimeframes, PairWithTimeframe
 from freqtrade.data.history import load_pair_history
-from freqtrade.enums import CandleType, RunMode
+from freqtrade.enums import CandleType, RPCMessageType, RunMode
 from freqtrade.exceptions import ExchangeError, OperationalException
 from freqtrade.exchange import Exchange, timeframe_to_seconds
+from freqtrade.rpc import RPCManager
 from freqtrade.util import PeriodicCache
 
 
@@ -28,17 +29,33 @@ MAX_DATAFRAME_CANDLES = 1000
 
 class DataProvider:
 
-    def __init__(self, config: Config, exchange: Optional[Exchange], pairlists=None) -> None:
+    def __init__(
+        self,
+        config: Config,
+        exchange: Optional[Exchange],
+        pairlists=None,
+        rpc: Optional[RPCManager] = None
+    ) -> None:
         self._config = config
         self._exchange = exchange
         self._pairlists = pairlists
+        self.__rpc = rpc
         self.__cached_pairs: Dict[PairWithTimeframe, Tuple[DataFrame, datetime]] = {}
         self.__slice_index: Optional[int] = None
         self.__cached_pairs_backtesting: Dict[PairWithTimeframe, DataFrame] = {}
+        self.__producer_pairs_df: Dict[str,
+                                       Dict[PairWithTimeframe, Tuple[DataFrame, datetime]]] = {}
+        self.__producer_pairs: Dict[str, List[str]] = {}
         self._msg_queue: deque = deque()
 
+        self._default_candle_type = self._config.get('candle_type_def', CandleType.SPOT)
+        self._default_timeframe = self._config.get('timeframe', '1h')
+
         self.__msg_cache = PeriodicCache(
-            maxsize=1000, ttl=timeframe_to_seconds(self._config.get('timeframe', '1h')))
+            maxsize=1000, ttl=timeframe_to_seconds(self._default_timeframe))
+
+        self.producers = self._config.get('external_message_consumer', {}).get('producers', [])
+        self.external_data_enabled = len(self.producers) > 0
 
     def _set_dataframe_max_index(self, limit_index: int):
         """
@@ -63,9 +80,110 @@ class DataProvider:
         :param dataframe: analyzed dataframe
         :param candle_type: Any of the enum CandleType (must match trading mode!)
         """
-        self.__cached_pairs[(pair, timeframe, candle_type)] = (
+        pair_key = (pair, timeframe, candle_type)
+        self.__cached_pairs[pair_key] = (
             dataframe, datetime.now(timezone.utc))
 
+    # For multiple producers we will want to merge the pairlists instead of overwriting
+    def _set_producer_pairs(self, pairlist: List[str], producer_name: str = "default"):
+        """
+        Set the pairs received to later be used.
+
+        :param pairlist: List of pairs
+        """
+        self.__producer_pairs[producer_name] = pairlist
+
+    def get_producer_pairs(self, producer_name: str = "default") -> List[str]:
+        """
+        Get the pairs cached from the producer
+
+        :returns: List of pairs
+        """
+        return self.__producer_pairs.get(producer_name, []).copy()
+
+    def _emit_df(
+        self,
+        pair_key: PairWithTimeframe,
+        dataframe: DataFrame
+    ) -> None:
+        """
+        Send this dataframe as an ANALYZED_DF message to RPC
+
+        :param pair_key: PairWithTimeframe tuple
+        :param data: Tuple containing the DataFrame and the datetime it was cached
+        """
+        if self.__rpc:
+            self.__rpc.send_msg(
+                {
+                    'type': RPCMessageType.ANALYZED_DF,
+                    'data': {
+                        'key': pair_key,
+                        'df': dataframe,
+                        'la': datetime.now(timezone.utc)
+                    }
+                }
+            )
+
+    def _add_external_df(
+        self,
+        pair: str,
+        dataframe: DataFrame,
+        last_analyzed: datetime,
+        timeframe: str,
+        candle_type: CandleType,
+        producer_name: str = "default"
+    ) -> None:
+        """
+        Add the pair data to this class from an external source.
+
+        :param pair: pair to get the data for
+        :param timeframe: Timeframe to get data for
+        :param candle_type: Any of the enum CandleType (must match trading mode!)
+        """
+        pair_key = (pair, timeframe, candle_type)
+
+        if producer_name not in self.__producer_pairs_df:
+            self.__producer_pairs_df[producer_name] = {}
+
+        _last_analyzed = datetime.now(timezone.utc) if not last_analyzed else last_analyzed
+
+        self.__producer_pairs_df[producer_name][pair_key] = (dataframe, _last_analyzed)
+        logger.debug(f"External DataFrame for {pair_key} from {producer_name} added.")
+
+    def get_producer_df(
+        self,
+        pair: str,
+        timeframe: Optional[str] = None,
+        candle_type: Optional[CandleType] = None,
+        producer_name: str = "default"
+    ) -> Tuple[DataFrame, datetime]:
+        """
+        Get the pair data from producers.
+
+        :param pair: pair to get the data for
+        :param timeframe: Timeframe to get data for
+        :param candle_type: Any of the enum CandleType (must match trading mode!)
+        :returns: Tuple of the DataFrame and last analyzed timestamp
+        """
+        _timeframe = self._default_timeframe if not timeframe else timeframe
+        _candle_type = self._default_candle_type if not candle_type else candle_type
+
+        pair_key = (pair, _timeframe, _candle_type)
+
+        # If we have no data from this Producer yet
+        if producer_name not in self.__producer_pairs_df:
+            # We don't have this data yet, return empty DataFrame and datetime (01-01-1970)
+            return (DataFrame(), datetime.fromtimestamp(0, tz=timezone.utc))
+
+        # If we do have data from that Producer, but no data on this pair_key
+        if pair_key not in self.__producer_pairs_df[producer_name]:
+            # We don't have this data yet, return empty DataFrame and datetime (01-01-1970)
+            return (DataFrame(), datetime.fromtimestamp(0, tz=timezone.utc))
+
+        # We have it, return this data
+        df, la = self.__producer_pairs_df[producer_name][pair_key]
+        return (df.copy(), la)
+
     def add_pairlisthandler(self, pairlists) -> None:
         """
         Allow adding pairlisthandler after initialization
@@ -90,8 +208,10 @@ class DataProvider:
         if saved_pair not in self.__cached_pairs_backtesting:
             timerange = TimeRange.parse_timerange(None if self._config.get(
                 'timerange') is None else str(self._config.get('timerange')))
-            # Move informative start time respecting startup_candle_count
+
-            startup_candles = self.get_required_startup(str(timeframe))
+            # It is not necessary to add the training candles, as they
+            # were already added at the beginning of the backtest.
+            startup_candles = self.get_required_startup(str(timeframe), False)
             tf_seconds = timeframe_to_seconds(str(timeframe))
             timerange.subtract_start(tf_seconds * startup_candles)
             self.__cached_pairs_backtesting[saved_pair] = load_pair_history(
@@ -105,7 +225,7 @@ class DataProvider:
             )
         return self.__cached_pairs_backtesting[saved_pair].copy()
 
-    def get_required_startup(self, timeframe: str) -> int:
+    def get_required_startup(self, timeframe: str, add_train_candles: bool = True) -> int:
         freqai_config = self._config.get('freqai', {})
         if not freqai_config.get('enabled', False):
             return self._config.get('startup_candle_count', 0)
@@ -115,7 +235,9 @@ class DataProvider:
             # make sure the startupcandles is at least the set maximum indicator periods
             self._config['startup_candle_count'] = max(startup_candles, max(indicator_periods))
             tf_seconds = timeframe_to_seconds(timeframe)
-            train_candles = freqai_config['train_period_days'] * 86400 / tf_seconds
+            train_candles = 0
+            if add_train_candles:
+                train_candles = freqai_config['train_period_days'] * 86400 / tf_seconds
             total_candles = int(self._config['startup_candle_count'] + train_candles)
             logger.info(f'Increasing startup_candle_count for freqai to {total_candles}')
             return total_candles

freqtrade/data/history/featherdatahandler.py

@@ -0,0 +1,130 @@
+import logging
+from typing import Optional
+
+from pandas import DataFrame, read_feather, to_datetime
+
+from freqtrade.configuration import TimeRange
+from freqtrade.constants import DEFAULT_DATAFRAME_COLUMNS, TradeList
+from freqtrade.enums import CandleType
+
+from .idatahandler import IDataHandler
+
+
+logger = logging.getLogger(__name__)
+
+
+class FeatherDataHandler(IDataHandler):
+
+    _columns = DEFAULT_DATAFRAME_COLUMNS
+
+    def ohlcv_store(
+            self, pair: str, timeframe: str, data: DataFrame, candle_type: CandleType) -> None:
+        """
+        Store data in json format "values".
+            format looks as follows:
+            [[<date>,<open>,<high>,<low>,<close>]]
+        :param pair: Pair - used to generate filename
+        :param timeframe: Timeframe - used to generate filename
+        :param data: Dataframe containing OHLCV data
+        :param candle_type: Any of the enum CandleType (must match trading mode!)
+        :return: None
+        """
+        filename = self._pair_data_filename(self._datadir, pair, timeframe, candle_type)
+        self.create_dir_if_needed(filename)
+
+        data.reset_index(drop=True).loc[:, self._columns].to_feather(
+            filename, compression_level=9, compression='lz4')
+
+    def _ohlcv_load(self, pair: str, timeframe: str,
+                    timerange: Optional[TimeRange], candle_type: CandleType
+                    ) -> DataFrame:
+        """
+        Internal method used to load data for one pair from disk.
+        Implements the loading and conversion to a Pandas dataframe.
+        Timerange trimming and dataframe validation happens outside of this method.
+        :param pair: Pair to load data
+        :param timeframe: Timeframe (e.g. "5m")
+        :param timerange: Limit data to be loaded to this timerange.
+                        Optionally implemented by subclasses to avoid loading
+                        all data where possible.
+        :param candle_type: Any of the enum CandleType (must match trading mode!)
+        :return: DataFrame with ohlcv data, or empty DataFrame
+        """
+        filename = self._pair_data_filename(
+            self._datadir, pair, timeframe, candle_type=candle_type)
+        if not filename.exists():
+            # Fallback mode for 1M files
+            filename = self._pair_data_filename(
+                self._datadir, pair, timeframe, candle_type=candle_type, no_timeframe_modify=True)
+            if not filename.exists():
+                return DataFrame(columns=self._columns)
+
+        pairdata = read_feather(filename)
+        pairdata.columns = self._columns
+        pairdata = pairdata.astype(dtype={'open': 'float', 'high': 'float',
+                                          'low': 'float', 'close': 'float', 'volume': 'float'})
+        pairdata['date'] = to_datetime(pairdata['date'],
+                                       unit='ms',
+                                       utc=True,
+                                       infer_datetime_format=True)
+        return pairdata
+
+    def ohlcv_append(
+        self,
+        pair: str,
+        timeframe: str,
+        data: DataFrame,
+        candle_type: CandleType
+    ) -> None:
+        """
+        Append data to existing data structures
+        :param pair: Pair
+        :param timeframe: Timeframe this ohlcv data is for
+        :param data: Data to append.
+        :param candle_type: Any of the enum CandleType (must match trading mode!)
+        """
+        raise NotImplementedError()
+
+    def trades_store(self, pair: str, data: TradeList) -> None:
+        """
+        Store trades data (list of Dicts) to file
+        :param pair: Pair - used for filename
+        :param data: List of Lists containing trade data,
+                     column sequence as in DEFAULT_TRADES_COLUMNS
+        """
+        # filename = self._pair_trades_filename(self._datadir, pair)
+
+        raise NotImplementedError()
+        # array = pa.array(data)
+        # array
+        # feather.write_feather(data, filename)
+
+    def trades_append(self, pair: str, data: TradeList):
+        """
+        Append data to existing files
+        :param pair: Pair - used for filename
+        :param data: List of Lists containing trade data,
+                     column sequence as in DEFAULT_TRADES_COLUMNS
+        """
+        raise NotImplementedError()
+
+    def _trades_load(self, pair: str, timerange: Optional[TimeRange] = None) -> TradeList:
+        """
+        Load a pair from file, either .json.gz or .json
+        # TODO: respect timerange ...
+        :param pair: Load trades for this pair
+        :param timerange: Timerange to load trades for - currently not implemented
+        :return: List of trades
+        """
+        raise NotImplementedError()
+        # filename = self._pair_trades_filename(self._datadir, pair)
+        # tradesdata = misc.file_load_json(filename)
+
+        # if not tradesdata:
+        #     return []
+
+        # return tradesdata
+
+    @classmethod
+    def _get_file_extension(cls):
+        return "feather"

freqtrade/data/history/hdf5datahandler.py

@@ -81,6 +81,7 @@ class HDF5DataHandler(IDataHandler):
             raise ValueError("Wrong dataframe format")
         pairdata = pairdata.astype(dtype={'open': 'float', 'high': 'float',
                                           'low': 'float', 'close': 'float', 'volume': 'float'})
+        pairdata = pairdata.reset_index(drop=True)
         return pairdata
 
     def ohlcv_append(

freqtrade/data/history/idatahandler.py

@@ -375,6 +375,12 @@ def get_datahandlerclass(datatype: str) -> Type[IDataHandler]:
     elif datatype == 'hdf5':
         from .hdf5datahandler import HDF5DataHandler
         return HDF5DataHandler
+    elif datatype == 'feather':
+        from .featherdatahandler import FeatherDataHandler
+        return FeatherDataHandler
+    elif datatype == 'parquet':
+        from .parquetdatahandler import ParquetDataHandler
+        return ParquetDataHandler
     else:
         raise ValueError(f"No datahandler for datatype {datatype} available.")
 

freqtrade/data/history/parquetdatahandler.py

@@ -0,0 +1,129 @@
+import logging
+from typing import Optional
+
+from pandas import DataFrame, read_parquet, to_datetime
+
+from freqtrade.configuration import TimeRange
+from freqtrade.constants import DEFAULT_DATAFRAME_COLUMNS, TradeList
+from freqtrade.enums import CandleType
+
+from .idatahandler import IDataHandler
+
+
+logger = logging.getLogger(__name__)
+
+
+class ParquetDataHandler(IDataHandler):
+
+    _columns = DEFAULT_DATAFRAME_COLUMNS
+
+    def ohlcv_store(
+            self, pair: str, timeframe: str, data: DataFrame, candle_type: CandleType) -> None:
+        """
+        Store data in json format "values".
+            format looks as follows:
+            [[<date>,<open>,<high>,<low>,<close>]]
+        :param pair: Pair - used to generate filename
+        :param timeframe: Timeframe - used to generate filename
+        :param data: Dataframe containing OHLCV data
+        :param candle_type: Any of the enum CandleType (must match trading mode!)
+        :return: None
+        """
+        filename = self._pair_data_filename(self._datadir, pair, timeframe, candle_type)
+        self.create_dir_if_needed(filename)
+
+        data.reset_index(drop=True).loc[:, self._columns].to_parquet(filename)
+
+    def _ohlcv_load(self, pair: str, timeframe: str,
+                    timerange: Optional[TimeRange], candle_type: CandleType
+                    ) -> DataFrame:
+        """
+        Internal method used to load data for one pair from disk.
+        Implements the loading and conversion to a Pandas dataframe.
+        Timerange trimming and dataframe validation happens outside of this method.
+        :param pair: Pair to load data
+        :param timeframe: Timeframe (e.g. "5m")
+        :param timerange: Limit data to be loaded to this timerange.
+                        Optionally implemented by subclasses to avoid loading
+                        all data where possible.
+        :param candle_type: Any of the enum CandleType (must match trading mode!)
+        :return: DataFrame with ohlcv data, or empty DataFrame
+        """
+        filename = self._pair_data_filename(
+            self._datadir, pair, timeframe, candle_type=candle_type)
+        if not filename.exists():
+            # Fallback mode for 1M files
+            filename = self._pair_data_filename(
+                self._datadir, pair, timeframe, candle_type=candle_type, no_timeframe_modify=True)
+            if not filename.exists():
+                return DataFrame(columns=self._columns)
+
+        pairdata = read_parquet(filename)
+        pairdata.columns = self._columns
+        pairdata = pairdata.astype(dtype={'open': 'float', 'high': 'float',
+                                          'low': 'float', 'close': 'float', 'volume': 'float'})
+        pairdata['date'] = to_datetime(pairdata['date'],
+                                       unit='ms',
+                                       utc=True,
+                                       infer_datetime_format=True)
+        return pairdata
+
+    def ohlcv_append(
+        self,
+        pair: str,
+        timeframe: str,
+        data: DataFrame,
+        candle_type: CandleType
+    ) -> None:
+        """
+        Append data to existing data structures
+        :param pair: Pair
+        :param timeframe: Timeframe this ohlcv data is for
+        :param data: Data to append.
+        :param candle_type: Any of the enum CandleType (must match trading mode!)
+        """
+        raise NotImplementedError()
+
+    def trades_store(self, pair: str, data: TradeList) -> None:
+        """
+        Store trades data (list of Dicts) to file
+        :param pair: Pair - used for filename
+        :param data: List of Lists containing trade data,
+                     column sequence as in DEFAULT_TRADES_COLUMNS
+        """
+        # filename = self._pair_trades_filename(self._datadir, pair)
+
+        raise NotImplementedError()
+        # array = pa.array(data)
+        # array
+        # feather.write_feather(data, filename)
+
+    def trades_append(self, pair: str, data: TradeList):
+        """
+        Append data to existing files
+        :param pair: Pair - used for filename
+        :param data: List of Lists containing trade data,
+                     column sequence as in DEFAULT_TRADES_COLUMNS
+        """
+        raise NotImplementedError()
+
+    def _trades_load(self, pair: str, timerange: Optional[TimeRange] = None) -> TradeList:
+        """
+        Load a pair from file, either .json.gz or .json
+        # TODO: respect timerange ...
+        :param pair: Load trades for this pair
+        :param timerange: Timerange to load trades for - currently not implemented
+        :return: List of trades
+        """
+        raise NotImplementedError()
+        # filename = self._pair_trades_filename(self._datadir, pair)
+        # tradesdata = misc.file_load_json(filename)
+
+        # if not tradesdata:
+        #     return []
+
+        # return tradesdata
+
+    @classmethod
+    def _get_file_extension(cls):
+        return "parquet"

freqtrade/enums/__init__.py

@@ -6,7 +6,7 @@ from freqtrade.enums.exittype import ExitType
 from freqtrade.enums.hyperoptstate import HyperoptState
 from freqtrade.enums.marginmode import MarginMode
 from freqtrade.enums.ordertypevalue import OrderTypeValues
-from freqtrade.enums.rpcmessagetype import RPCMessageType
+from freqtrade.enums.rpcmessagetype import RPCMessageType, RPCRequestType
 from freqtrade.enums.runmode import NON_UTIL_MODES, OPTIMIZE_MODES, TRADING_MODES, RunMode
 from freqtrade.enums.signaltype import SignalDirection, SignalTagType, SignalType
 from freqtrade.enums.state import State

freqtrade/enums/rpcmessagetype.py

@@ -1,7 +1,7 @@
 from enum import Enum
 
 
-class RPCMessageType(Enum):
+class RPCMessageType(str, Enum):
     STATUS = 'status'
     WARNING = 'warning'
     STARTUP = 'startup'
@@ -19,8 +19,19 @@ class RPCMessageType(Enum):
 
     STRATEGY_MSG = 'strategy_msg'
 
+    WHITELIST = 'whitelist'
+    ANALYZED_DF = 'analyzed_df'
+
     def __repr__(self):
         return self.value
 
     def __str__(self):
         return self.value
+
+
+# Enum for parsing requests from ws consumers
+class RPCRequestType(str, Enum):
+    SUBSCRIBE = 'subscribe'
+
+    WHITELIST = 'whitelist'
+    ANALYZED_DF = 'analyzed_df'

freqtrade/exchange/exchange.py

@@ -2891,7 +2891,7 @@ def amount_to_contracts(amount: float, contract_size: Optional[float]) -> float:
     :return: num-contracts
     """
     if contract_size and contract_size != 1:
-        return amount / contract_size
+        return float(FtPrecise(amount) / FtPrecise(contract_size))
     else:
         return amount
 
@@ -2905,7 +2905,7 @@ def contracts_to_amount(num_contracts: float, contract_size: Optional[float]) ->
     """
 
     if contract_size and contract_size != 1:
-        return num_contracts * contract_size
+        return float(FtPrecise(num_contracts) * FtPrecise(contract_size))
     else:
         return num_contracts
 

freqtrade/freqai/base_models/BaseClassifierModel.py

@@ -1,4 +1,5 @@
 import logging
+from time import time
 from typing import Any, Tuple
 
 import numpy as np
@@ -32,7 +33,9 @@ class BaseClassifierModel(IFreqaiModel):
         :model: Trained model which can be used to inference (self.predict)
         """
 
-        logger.info("-------------------- Starting training " f"{pair} --------------------")
+        logger.info(f"-------------------- Starting training {pair} --------------------")
+
+        start_time = time()
 
         # filter the features requested by user in the configuration file and elegantly handle NaNs
         features_filtered, labels_filtered = dk.filter_features(
@@ -45,10 +48,10 @@ class BaseClassifierModel(IFreqaiModel):
         start_date = unfiltered_df["date"].iloc[0].strftime("%Y-%m-%d")
         end_date = unfiltered_df["date"].iloc[-1].strftime("%Y-%m-%d")
         logger.info(f"-------------------- Training on data from {start_date} to "
-                    f"{end_date}--------------------")
+                    f"{end_date} --------------------")
         # split data into train/test data.
         data_dictionary = dk.make_train_test_datasets(features_filtered, labels_filtered)
-        if not self.freqai_info.get('fit_live_predictions', 0) or not self.live:
+        if not self.freqai_info.get("fit_live_predictions", 0) or not self.live:
             dk.fit_labels()
         # normalize all data based on train_dataset only
         data_dictionary = dk.normalize_data(data_dictionary)
@@ -57,13 +60,16 @@ class BaseClassifierModel(IFreqaiModel):
         self.data_cleaning_train(dk)
 
         logger.info(
-            f'Training model on {len(dk.data_dictionary["train_features"].columns)}' " features"
+            f"Training model on {len(dk.data_dictionary['train_features'].columns)} features"
         )
-        logger.info(f'Training model on {len(data_dictionary["train_features"])} data points')
+        logger.info(f"Training model on {len(data_dictionary['train_features'])} data points")
 
         model = self.fit(data_dictionary, dk)
 
-        logger.info(f"--------------------done training {pair}--------------------")
+        end_time = time()
+
+        logger.info(f"-------------------- Done training {pair} "
+                    f"({end_time - start_time:.2f} secs) --------------------")
 
         return model
 

freqtrade/freqai/base_models/BaseRegressionModel.py

@@ -1,4 +1,5 @@
 import logging
+from time import time
 from typing import Any, Tuple
 
 import numpy as np
@@ -31,7 +32,9 @@ class BaseRegressionModel(IFreqaiModel):
         :model: Trained model which can be used to inference (self.predict)
         """
 
-        logger.info("-------------------- Starting training " f"{pair} --------------------")
+        logger.info(f"-------------------- Starting training {pair} --------------------")
+
+        start_time = time()
 
         # filter the features requested by user in the configuration file and elegantly handle NaNs
         features_filtered, labels_filtered = dk.filter_features(
@@ -44,10 +47,10 @@ class BaseRegressionModel(IFreqaiModel):
         start_date = unfiltered_df["date"].iloc[0].strftime("%Y-%m-%d")
         end_date = unfiltered_df["date"].iloc[-1].strftime("%Y-%m-%d")
         logger.info(f"-------------------- Training on data from {start_date} to "
-                    f"{end_date}--------------------")
+                    f"{end_date} --------------------")
         # split data into train/test data.
         data_dictionary = dk.make_train_test_datasets(features_filtered, labels_filtered)
-        if not self.freqai_info.get('fit_live_predictions', 0) or not self.live:
+        if not self.freqai_info.get("fit_live_predictions", 0) or not self.live:
             dk.fit_labels()
         # normalize all data based on train_dataset only
         data_dictionary = dk.normalize_data(data_dictionary)
@@ -56,13 +59,16 @@ class BaseRegressionModel(IFreqaiModel):
         self.data_cleaning_train(dk)
 
         logger.info(
-            f'Training model on {len(dk.data_dictionary["train_features"].columns)}' " features"
+            f"Training model on {len(dk.data_dictionary['train_features'].columns)} features"
         )
-        logger.info(f'Training model on {len(data_dictionary["train_features"])} data points')
+        logger.info(f"Training model on {len(data_dictionary['train_features'])} data points")
 
         model = self.fit(data_dictionary, dk)
 
-        logger.info(f"--------------------done training {pair}--------------------")
+        end_time = time()
+
+        logger.info(f"-------------------- Done training {pair} "
+                    f"({end_time - start_time:.2f} secs) --------------------")
 
         return model
 

freqtrade/freqai/base_models/BaseTensorFlowModel.py

@@ -1,4 +1,5 @@
 import logging
+from time import time
 from typing import Any
 
 from pandas import DataFrame
@@ -28,7 +29,9 @@ class BaseTensorFlowModel(IFreqaiModel):
         :model: Trained model which can be used to inference (self.predict)
         """
 
-        logger.info("-------------------- Starting training " f"{pair} --------------------")
+        logger.info(f"-------------------- Starting training {pair} --------------------")
+
+        start_time = time()
 
         # filter the features requested by user in the configuration file and elegantly handle NaNs
         features_filtered, labels_filtered = dk.filter_features(
@@ -41,10 +44,10 @@ class BaseTensorFlowModel(IFreqaiModel):
         start_date = unfiltered_df["date"].iloc[0].strftime("%Y-%m-%d")
         end_date = unfiltered_df["date"].iloc[-1].strftime("%Y-%m-%d")
         logger.info(f"-------------------- Training on data from {start_date} to "
-                    f"{end_date}--------------------")
+                    f"{end_date} --------------------")
         # split data into train/test data.
         data_dictionary = dk.make_train_test_datasets(features_filtered, labels_filtered)
-        if not self.freqai_info.get('fit_live_predictions', 0) or not self.live:
+        if not self.freqai_info.get("fit_live_predictions", 0) or not self.live:
             dk.fit_labels()
         # normalize all data based on train_dataset only
         data_dictionary = dk.normalize_data(data_dictionary)
@@ -53,12 +56,15 @@ class BaseTensorFlowModel(IFreqaiModel):
         self.data_cleaning_train(dk)
 
         logger.info(
-            f'Training model on {len(dk.data_dictionary["train_features"].columns)}' " features"
+            f"Training model on {len(dk.data_dictionary['train_features'].columns)} features"
         )
-        logger.info(f'Training model on {len(data_dictionary["train_features"])} data points')
+        logger.info(f"Training model on {len(data_dictionary['train_features'])} data points")
 
         model = self.fit(data_dictionary, dk)
 
-        logger.info(f"--------------------done training {pair}--------------------")
+        end_time = time()
+
+        logger.info(f"-------------------- Done training {pair} "
+                    f"({end_time - start_time:.2f} secs) --------------------")
 
         return model

freqtrade/freqai/base_models/FreqaiMultiOutputRegressor.py

@@ -1,4 +1,3 @@
-
 from joblib import Parallel
 from sklearn.multioutput import MultiOutputRegressor, _fit_estimator
 from sklearn.utils.fixes import delayed

freqtrade/freqai/data_drawer.py

@@ -313,6 +313,7 @@ class FreqaiDataDrawer:
         """
 
         dk.find_features(dataframe)
+        dk.find_labels(dataframe)
 
         full_labels = dk.label_list + dk.unique_class_list
 
@@ -376,7 +377,27 @@ class FreqaiDataDrawer:
         if self.config.get("freqai", {}).get("purge_old_models", False):
             self.purge_old_models()
 
-    # Functions pulled back from FreqaiDataKitchen because they relied on DataDrawer
+    def save_metadata(self, dk: FreqaiDataKitchen) -> None:
+        """
+        Saves only metadata for backtesting studies if user prefers
+        not to save model data. This saves tremendous amounts of space
+        for users generating huge studies.
+        This is only active when `save_backtest_models`: false (not default)
+        """
+        if not dk.data_path.is_dir():
+            dk.data_path.mkdir(parents=True, exist_ok=True)
+
+        save_path = Path(dk.data_path)
+
+        dk.data["data_path"] = str(dk.data_path)
+        dk.data["model_filename"] = str(dk.model_filename)
+        dk.data["training_features_list"] = list(dk.data_dictionary["train_features"].columns)
+        dk.data["label_list"] = dk.label_list
+
+        with open(save_path / f"{dk.model_filename}_metadata.json", "w") as fp:
+            rapidjson.dump(dk.data, fp, default=self.np_encoder, number_mode=rapidjson.NM_NATIVE)
+
+        return
 
     def save_data(self, model: Any, coin: str, dk: FreqaiDataKitchen) -> None:
         """
@@ -430,6 +451,16 @@ class FreqaiDataDrawer:
 
         return
 
+    def load_metadata(self, dk: FreqaiDataKitchen) -> None:
+        """
+        Load only metadata into datakitchen to increase performance during
+        presaved backtesting (prediction file loading).
+        """
+        with open(dk.data_path / f"{dk.model_filename}_metadata.json", "r") as fp:
+            dk.data = json.load(fp)
+            dk.training_features_list = dk.data["training_features_list"]
+            dk.label_list = dk.data["label_list"]
+
     def load_data(self, coin: str, dk: FreqaiDataKitchen) -> Any:
         """
         loads all data required to make a prediction on a sub-train time range

freqtrade/freqai/data_kitchen.py

@@ -466,27 +466,6 @@ class FreqaiDataKitchen:
 
         return df
 
-    def remove_training_from_backtesting(
-        self
-    ) -> DataFrame:
-        """
-        Function which takes the backtesting time range and
-        remove training data from dataframe, keeping only the
-        startup_candle_count candles
-        """
-        startup_candle_count = self.config.get('startup_candle_count', 0)
-        tf = self.config['timeframe']
-        tr = self.config["timerange"]
-
-        backtesting_timerange = TimeRange.parse_timerange(tr)
-        if startup_candle_count > 0 and backtesting_timerange:
-            backtesting_timerange.subtract_start(timeframe_to_seconds(tf) * startup_candle_count)
-
-        start = datetime.fromtimestamp(backtesting_timerange.startts, tz=timezone.utc)
-        df = self.return_dataframe
-        df = df.loc[df["date"] >= start, :]
-        return df
-
     def principal_component_analysis(self) -> None:
         """
         Performs Principal Component Analysis on the data for dimensionality reduction
@@ -775,12 +754,22 @@ class FreqaiDataKitchen:
 
     def compute_inlier_metric(self, set_='train') -> None:
         """
-
         Compute inlier metric from backwards distance distributions.
         This metric defines how well features from a timepoint fit
         into previous timepoints.
         """
 
+        def normalise(dataframe: DataFrame, key: str) -> DataFrame:
+            if set_ == 'train':
+                min_value = dataframe.min()
+                max_value = dataframe.max()
+                self.data[f'{key}_min'] = min_value
+                self.data[f'{key}_max'] = max_value
+            else:
+                min_value = self.data[f'{key}_min']
+                max_value = self.data[f'{key}_max']
+            return (dataframe - min_value) / (max_value - min_value)
+
         no_prev_pts = self.freqai_config["feature_parameters"]["inlier_metric_window"]
 
         if set_ == 'train':
@@ -825,7 +814,12 @@ class FreqaiDataKitchen:
         inliers = pd.DataFrame(index=distances.index)
         for key in distances.keys():
             current_distances = distances[key].dropna()
-            fit_params = stats.weibull_min.fit(current_distances)
+            current_distances = normalise(current_distances, key)
+            if set_ == 'train':
+                fit_params = stats.weibull_min.fit(current_distances)
+                self.data[f'{key}_fit_params'] = fit_params
+            else:
+                fit_params = self.data[f'{key}_fit_params']
             quantiles = stats.weibull_min.cdf(current_distances, *fit_params)
 
             df_inlier = pd.DataFrame(
@@ -837,7 +831,7 @@ class FreqaiDataKitchen:
 
         inlier_metric = pd.DataFrame(
             data=inliers.sum(axis=1) / no_prev_pts,
-            columns=['inlier_metric'],
+            columns=['%-inlier_metric'],
             index=compute_df.index
         )
 
@@ -887,11 +881,14 @@ class FreqaiDataKitchen:
         """
         column_names = dataframe.columns
         features = [c for c in column_names if "%" in c]
-        labels = [c for c in column_names if "&" in c]
         if not features:
             raise OperationalException("Could not find any features!")
 
         self.training_features_list = features
+
+    def find_labels(self, dataframe: DataFrame) -> None:
+        column_names = dataframe.columns
+        labels = [c for c in column_names if "&" in c]
         self.label_list = labels
 
     def check_if_pred_in_training_spaces(self) -> None:
@@ -979,8 +976,6 @@ class FreqaiDataKitchen:
 
         to_keep = [col for col in dataframe.columns if not col.startswith("&")]
         self.return_dataframe = pd.concat([dataframe[to_keep], self.full_df], axis=1)
-
-        self.return_dataframe = self.remove_training_from_backtesting()
         self.full_df = DataFrame()
 
         return
@@ -1214,7 +1209,8 @@ class FreqaiDataKitchen:
 
     def get_unique_classes_from_labels(self, dataframe: DataFrame) -> None:
 
-        self.find_features(dataframe)
+        # self.find_features(dataframe)
+        self.find_labels(dataframe)
 
         for key in self.label_list:
             if dataframe[key].dtype == object:

freqtrade/freqai/freqai_interface.py

@@ -65,7 +65,7 @@ class IFreqaiModel(ABC):
         self.first = True
         self.set_full_path()
         self.follow_mode: bool = self.freqai_info.get("follow_mode", False)
-        self.save_backtest_models: bool = self.freqai_info.get("save_backtest_models", False)
+        self.save_backtest_models: bool = self.freqai_info.get("save_backtest_models", True)
         if self.save_backtest_models:
             logger.info('Backtesting module configured to save all models.')
         self.dd = FreqaiDataDrawer(Path(self.full_path), self.config, self.follow_mode)
@@ -92,6 +92,7 @@ class IFreqaiModel(ABC):
         self.begin_time_train: float = 0
         self.base_tf_seconds = timeframe_to_seconds(self.config['timeframe'])
         self.continual_learning = self.freqai_info.get('continual_learning', False)
+        self.plot_features = self.ft_params.get("plot_feature_importances", 0)
 
         self._threads: List[threading.Thread] = []
         self._stop_event = threading.Event()
@@ -244,7 +245,8 @@ class IFreqaiModel(ABC):
         # following tr_train. Both of these windows slide through the
         # entire backtest
         for tr_train, tr_backtest in zip(dk.training_timeranges, dk.backtesting_timeranges):
-            (_, _, _) = self.dd.get_pair_dict_info(metadata["pair"])
+            pair = metadata["pair"]
+            (_, _, _) = self.dd.get_pair_dict_info(pair)
             train_it += 1
             total_trains = len(dk.backtesting_timeranges)
             self.training_timerange = tr_train
@@ -259,40 +261,42 @@ class IFreqaiModel(ABC):
                                                 tr_train.stopts,
                                                 tz=timezone.utc).strftime(DATETIME_PRINT_FORMAT)
             logger.info(
-                f"Training {metadata['pair']}, {self.pair_it}/{self.total_pairs} pairs"
+                f"Training {pair}, {self.pair_it}/{self.total_pairs} pairs"
                 f" from {tr_train_startts_str} to {tr_train_stopts_str}, {train_it}/{total_trains} "
                 "trains"
             )
 
             trained_timestamp_int = int(trained_timestamp.stopts)
             dk.data_path = Path(
-                dk.full_path
+                dk.full_path / f"sub-train-{pair.split('/')[0]}_{trained_timestamp_int}"
-                /
-                f"sub-train-{metadata['pair'].split('/')[0]}_{trained_timestamp_int}"
                 )
 
-            dk.set_new_model_names(metadata["pair"], trained_timestamp)
+            dk.set_new_model_names(pair, trained_timestamp)
 
             if dk.check_if_backtest_prediction_exists():
+                self.dd.load_metadata(dk)
+                self.check_if_feature_list_matches_strategy(dataframe_train, dk)
                 append_df = dk.get_backtesting_prediction()
                 dk.append_predictions(append_df)
             else:
-                if not self.model_exists(
+                if not self.model_exists(dk):
-                    metadata["pair"], dk, trained_timestamp=trained_timestamp_int
-                ):
                     dk.find_features(dataframe_train)
-                    self.model = self.train(dataframe_train, metadata["pair"], dk)
+                    dk.find_labels(dataframe_train)
-                    self.dd.pair_dict[metadata["pair"]]["trained_timestamp"] = int(
+                    self.model = self.train(dataframe_train, pair, dk)
+                    self.dd.pair_dict[pair]["trained_timestamp"] = int(
                         trained_timestamp.stopts)
-
+                    if self.plot_features:
+                        plot_feature_importance(self.model, pair, dk, self.plot_features)
                     if self.save_backtest_models:
                         logger.info('Saving backtest model to disk.')
-                        self.dd.save_data(self.model, metadata["pair"], dk)
+                        self.dd.save_data(self.model, pair, dk)
+                    else:
+                        logger.info('Saving metadata to disk.')
+                        self.dd.save_metadata(dk)
                 else:
-                    self.model = self.dd.load_data(metadata["pair"], dk)
+                    self.model = self.dd.load_data(pair, dk)
-
-                self.check_if_feature_list_matches_strategy(dataframe_train, dk)
 
+                # self.check_if_feature_list_matches_strategy(dataframe_train, dk)
                 pred_df, do_preds = self.predict(dataframe_backtest, dk)
                 append_df = dk.get_predictions_to_append(pred_df, do_preds)
                 dk.append_predictions(append_df)
@@ -371,8 +375,7 @@ class IFreqaiModel(ABC):
             self.dd.return_null_values_to_strategy(dataframe, dk)
             return dk
 
-        # ensure user is feeding the correct indicators to the model
+        dk.find_labels(dataframe)
-        self.check_if_feature_list_matches_strategy(dataframe, dk)
 
         self.build_strategy_return_arrays(dataframe, dk, metadata["pair"], trained_timestamp)
 
@@ -430,14 +433,16 @@ class IFreqaiModel(ABC):
         if "training_features_list_raw" in dk.data:
             feature_list = dk.data["training_features_list_raw"]
         else:
-            feature_list = dk.training_features_list
+            feature_list = dk.data['training_features_list']
         if dk.training_features_list != feature_list:
             raise OperationalException(
                 "Trying to access pretrained model with `identifier` "
                 "but found different features furnished by current strategy."
                 "Change `identifier` to train from scratch, or ensure the"
                 "strategy is furnishing the same features as the pretrained"
-                "model"
+                "model. In case of --strategy-list, please be aware that FreqAI "
+                "requires all strategies to maintain identical "
+                "populate_any_indicator() functions"
             )
 
     def data_cleaning_train(self, dk: FreqaiDataKitchen) -> None:
@@ -489,7 +494,7 @@ class IFreqaiModel(ABC):
         if ft_params.get(
             "principal_component_analysis", False
         ):
-            dk.pca_transform(self.dk.data_dictionary['prediction_features'])
+            dk.pca_transform(dk.data_dictionary['prediction_features'])
 
         if ft_params.get("use_SVM_to_remove_outliers", False):
             dk.use_SVM_to_remove_outliers(predict=True)
@@ -500,14 +505,10 @@ class IFreqaiModel(ABC):
         if ft_params.get("use_DBSCAN_to_remove_outliers", False):
             dk.use_DBSCAN_to_remove_outliers(predict=True)
 
-    def model_exists(
+        # ensure user is feeding the correct indicators to the model
-        self,
+        self.check_if_feature_list_matches_strategy(dk.data_dictionary['prediction_features'], dk)
-        pair: str,
+
-        dk: FreqaiDataKitchen,
+    def model_exists(self, dk: FreqaiDataKitchen) -> bool:
-        trained_timestamp: int = None,
-        model_filename: str = "",
-        scanning: bool = False,
-    ) -> bool:
         """
         Given a pair and path, check if a model already exists
         :param pair: pair e.g. BTC/USD
@@ -515,11 +516,11 @@ class IFreqaiModel(ABC):
         :return:
         :boolean: whether the model file exists or not.
         """
-        path_to_modelfile = Path(dk.data_path / f"{model_filename}_model.joblib")
+        path_to_modelfile = Path(dk.data_path / f"{dk.model_filename}_model.joblib")
         file_exists = path_to_modelfile.is_file()
-        if file_exists and not scanning:
+        if file_exists:
             logger.info("Found model at %s", dk.data_path / dk.model_filename)
-        elif not scanning:
+        else:
             logger.info("Could not find model at %s", dk.data_path / dk.model_filename)
         return file_exists
 
@@ -566,6 +567,7 @@ class IFreqaiModel(ABC):
 
         # find the features indicated by strategy and store in datakitchen
         dk.find_features(unfiltered_dataframe)
+        dk.find_labels(unfiltered_dataframe)
 
         model = self.train(unfiltered_dataframe, pair, dk)
 
@@ -573,8 +575,8 @@ class IFreqaiModel(ABC):
         dk.set_new_model_names(pair, new_trained_timerange)
         self.dd.save_data(model, pair, dk)
 
-        if self.freqai_info["feature_parameters"].get("plot_feature_importance", False):
+        if self.plot_features:
-            plot_feature_importance(model, pair, dk)
+            plot_feature_importance(model, pair, dk, self.plot_features)
 
         if self.freqai_info.get("purge_old_models", False):
             self.dd.purge_old_models()

freqtrade/freqai/utils.py

@@ -170,7 +170,7 @@ def plot_feature_importance(model: Any, pair: str, dk: FreqaiDataKitchen,
 
         # Data preparation
         fi_df = pd.DataFrame({
-            "feature_names": np.array(dk.training_features_list),
+            "feature_names": np.array(dk.data_dictionary['train_features'].columns),
             "feature_importance": np.array(feature_importance)
         })
         fi_df_top = fi_df.nlargest(count_max, "feature_importance")[::-1]

freqtrade/freqtradebot.py

@@ -29,6 +29,7 @@ from freqtrade.plugins.pairlistmanager import PairListManager
 from freqtrade.plugins.protectionmanager import ProtectionManager
 from freqtrade.resolvers import ExchangeResolver, StrategyResolver
 from freqtrade.rpc import RPCManager
+from freqtrade.rpc.external_message_consumer import ExternalMessageConsumer
 from freqtrade.strategy.interface import IStrategy
 from freqtrade.strategy.strategy_wrapper import strategy_safe_wrapper
 from freqtrade.util import FtPrecise
@@ -72,6 +73,8 @@ class FreqtradeBot(LoggingMixin):
 
         PairLocks.timeframe = self.config['timeframe']
 
+        self.pairlists = PairListManager(self.exchange, self.config)
+
         # RPC runs in separate threads, can start handling external commands just after
         # initialization, even before Freqtradebot has a chance to start its throttling,
         # so anything in the Freqtradebot instance should be ready (initialized), including
@@ -79,9 +82,7 @@ class FreqtradeBot(LoggingMixin):
         # Keep this at the end of this initialization method.
         self.rpc: RPCManager = RPCManager(self)
 
-        self.pairlists = PairListManager(self.exchange, self.config)
+        self.dataprovider = DataProvider(self.config, self.exchange, self.pairlists, self.rpc)
-
-        self.dataprovider = DataProvider(self.config, self.exchange, self.pairlists)
 
         # Attach Dataprovider to strategy instance
         self.strategy.dp = self.dataprovider
@@ -92,6 +93,10 @@ class FreqtradeBot(LoggingMixin):
         self.edge = Edge(self.config, self.exchange, self.strategy) if \
             self.config.get('edge', {}).get('enabled', False) else None
 
+        # Init ExternalMessageConsumer if enabled
+        self.emc = ExternalMessageConsumer(self.config, self.dataprovider) if \
+            self.config.get('external_message_consumer', {}).get('enabled', False) else None
+
         self.active_pair_whitelist = self._refresh_active_whitelist()
 
         # Set initial bot state from config
@@ -151,9 +156,11 @@ class FreqtradeBot(LoggingMixin):
         finally:
             self.strategy.ft_bot_cleanup()
 
-            self.rpc.cleanup()
+        self.rpc.cleanup()
-            Trade.commit()
+        if self.emc:
-            self.exchange.close()
+            self.emc.shutdown()
+        Trade.commit()
+        self.exchange.close()
 
     def startup(self) -> None:
         """
@@ -254,6 +261,7 @@ class FreqtradeBot(LoggingMixin):
         pairs that have open trades.
         """
         # Refresh whitelist
+        _prev_whitelist = self.pairlists.whitelist
         self.pairlists.refresh_pairlist()
         _whitelist = self.pairlists.whitelist
 
@@ -266,6 +274,11 @@ class FreqtradeBot(LoggingMixin):
             # Extend active-pair whitelist with pairs of open trades
             # It ensures that candle (OHLCV) data are downloaded for open trades as well
             _whitelist.extend([trade.pair for trade in trades if trade.pair not in _whitelist])
+
+        # Called last to include the included pairs
+        if _prev_whitelist != _whitelist:
+            self.rpc.send_msg({'type': RPCMessageType.WHITELIST, 'data': _whitelist})
+
         return _whitelist
 
     def get_free_open_trades(self) -> int:

freqtrade/misc.py

@@ -10,9 +10,11 @@ from typing import Any, Iterator, List
 from typing.io import IO
 from urllib.parse import urlparse
 
+import pandas
 import rapidjson
 
 from freqtrade.constants import DECIMAL_PER_COIN_FALLBACK, DECIMALS_PER_COIN
+from freqtrade.enums import SignalTagType, SignalType
 
 
 logger = logging.getLogger(__name__)
@@ -249,3 +251,41 @@ def parse_db_uri_for_logging(uri: str):
         return uri
     pwd = parsed_db_uri.netloc.split(':')[1].split('@')[0]
     return parsed_db_uri.geturl().replace(f':{pwd}@', ':*****@')
+
+
+def dataframe_to_json(dataframe: pandas.DataFrame) -> str:
+    """
+    Serialize a DataFrame for transmission over the wire using JSON
+    :param dataframe: A pandas DataFrame
+    :returns: A JSON string of the pandas DataFrame
+    """
+    return dataframe.to_json(orient='split')
+
+
+def json_to_dataframe(data: str) -> pandas.DataFrame:
+    """
+    Deserialize JSON into a DataFrame
+    :param data: A JSON string
+    :returns: A pandas DataFrame from the JSON string
+    """
+    dataframe = pandas.read_json(data, orient='split')
+    if 'date' in dataframe.columns:
+        dataframe['date'] = pandas.to_datetime(dataframe['date'], unit='ms', utc=True)
+
+    return dataframe
+
+
+def remove_entry_exit_signals(dataframe: pandas.DataFrame):
+    """
+    Remove Entry and Exit signals from a DataFrame
+
+    :param dataframe: The DataFrame to remove signals from
+    """
+    dataframe[SignalType.ENTER_LONG.value] = 0
+    dataframe[SignalType.EXIT_LONG.value] = 0
+    dataframe[SignalType.ENTER_SHORT.value] = 0
+    dataframe[SignalType.EXIT_SHORT.value] = 0
+    dataframe[SignalTagType.ENTER_TAG.value] = None
+    dataframe[SignalTagType.EXIT_TAG.value] = None
+
+    return dataframe

freqtrade/optimize/backtesting.py

@@ -91,8 +91,8 @@ class Backtesting:
 
         if self.config.get('strategy_list'):
             if self.config.get('freqai', {}).get('enabled', False):
-                raise OperationalException(
+                logger.warning("Using --strategy-list with FreqAI REQUIRES all strategies "
-                    "You can't use strategy_list and freqai at the same time.")
+                               "to have identical populate_any_indicators.")
             for strat in list(self.config['strategy_list']):
                 stratconf = deepcopy(self.config)
                 stratconf['strategy'] = strat
@@ -139,9 +139,14 @@ class Backtesting:
 
         # Get maximum required startup period
         self.required_startup = max([strat.startup_candle_count for strat in self.strategylist])
+        self.exchange.validate_required_startup_candles(self.required_startup, self.timeframe)
+
+        if self.config.get('freqai', {}).get('enabled', False):
+            # For FreqAI, increase the required_startup to includes the training data
+            self.required_startup = self.dataprovider.get_required_startup(self.timeframe)
+
         # Add maximum startup candle count to configuration for informative pairs support
         self.config['startup_candle_count'] = self.required_startup
-        self.exchange.validate_required_startup_candles(self.required_startup, self.timeframe)
 
         self.trading_mode: TradingMode = config.get('trading_mode', TradingMode.SPOT)
         # strategies which define "can_short=True" will fail to load in Spot mode.
@@ -217,7 +222,7 @@ class Backtesting:
             pairs=self.pairlists.whitelist,
             timeframe=self.timeframe,
             timerange=self.timerange,
-            startup_candles=self.dataprovider.get_required_startup(self.timeframe),
+            startup_candles=self.config['startup_candle_count'],
             fail_without_data=True,
             data_format=self.config.get('dataformat_ohlcv', 'json'),
             candle_type=self.config.get('candle_type_def', CandleType.SPOT)

freqtrade/rpc/api_server/api_auth.py

@@ -1,8 +1,10 @@
+import logging
 import secrets
 from datetime import datetime, timedelta
+from typing import Any, Dict, Union
 
 import jwt
-from fastapi import APIRouter, Depends, HTTPException, status
+from fastapi import APIRouter, Depends, HTTPException, Query, WebSocket, status
 from fastapi.security import OAuth2PasswordBearer
 from fastapi.security.http import HTTPBasic, HTTPBasicCredentials
 
@@ -10,6 +12,8 @@ from freqtrade.rpc.api_server.api_schemas import AccessAndRefreshToken, AccessTo
 from freqtrade.rpc.api_server.deps import get_api_config
 
 
+logger = logging.getLogger(__name__)
+
 ALGORITHM = "HS256"
 
 router_login = APIRouter()
@@ -25,7 +29,7 @@ httpbasic = HTTPBasic(auto_error=False)
 oauth2_scheme = OAuth2PasswordBearer(tokenUrl="token", auto_error=False)
 
 
-def get_user_from_token(token, secret_key: str, token_type: str = "access"):
+def get_user_from_token(token, secret_key: str, token_type: str = "access") -> str:
     credentials_exception = HTTPException(
         status_code=status.HTTP_401_UNAUTHORIZED,
         detail="Could not validate credentials",
@@ -44,6 +48,45 @@ def get_user_from_token(token, secret_key: str, token_type: str = "access"):
     return username
 
 
+# This should be reimplemented to better realign with the existing tools provided
+# by FastAPI regarding API Tokens
+# https://github.com/tiangolo/fastapi/blob/master/fastapi/security/api_key.py
+async def validate_ws_token(
+    ws: WebSocket,
+    ws_token: Union[str, None] = Query(default=None, alias="token"),
+    api_config: Dict[str, Any] = Depends(get_api_config)
+):
+    secret_ws_token = api_config.get('ws_token', None)
+    secret_jwt_key = api_config.get('jwt_secret_key', 'super-secret')
+
+    # Check if ws_token is/in secret_ws_token
+    if ws_token and secret_ws_token:
+        is_valid_ws_token = False
+        if isinstance(secret_ws_token, str):
+            is_valid_ws_token = secrets.compare_digest(secret_ws_token, ws_token)
+        elif isinstance(secret_ws_token, list):
+            is_valid_ws_token = any([
+                secrets.compare_digest(potential, ws_token)
+                for potential in secret_ws_token
+            ])
+
+        if is_valid_ws_token:
+            return ws_token
+
+    # Check if ws_token is a JWT
+    try:
+        user = get_user_from_token(ws_token, secret_jwt_key)
+        return user
+    # If the token is a jwt, and it's valid return the user
+    except HTTPException:
+        pass
+
+    # No checks passed, deny the connection
+    logger.debug("Denying websocket request.")
+    # If it doesn't match, close the websocket connection
+    await ws.close(code=status.WS_1008_POLICY_VIOLATION)
+
+
 def create_token(data: dict, secret_key: str, token_type: str = "access") -> str:
     to_encode = data.copy()
     if token_type == "access":

freqtrade/rpc/api_server/api_v1.py

@@ -38,7 +38,8 @@ logger = logging.getLogger(__name__)
 # 2.15: Add backtest history endpoints
 # 2.16: Additional daily metrics
 # 2.17: Forceentry - leverage, partial force_exit
-API_VERSION = 2.17
+# 2.20: Add websocket endpoints
+API_VERSION = 2.20
 
 # Public API, requires no auth.
 router_public = APIRouter()

freqtrade/rpc/api_server/api_ws.py

@@ -0,0 +1,140 @@
+import logging
+from typing import Any, Dict
+
+from fastapi import APIRouter, Depends, WebSocketDisconnect
+from fastapi.websockets import WebSocket, WebSocketState
+from pydantic import ValidationError
+
+from freqtrade.enums import RPCMessageType, RPCRequestType
+from freqtrade.rpc.api_server.api_auth import validate_ws_token
+from freqtrade.rpc.api_server.deps import get_channel_manager, get_rpc
+from freqtrade.rpc.api_server.ws import WebSocketChannel
+from freqtrade.rpc.api_server.ws_schemas import (WSAnalyzedDFMessage, WSMessageSchema,
+                                                 WSRequestSchema, WSWhitelistMessage)
+from freqtrade.rpc.rpc import RPC
+
+
+logger = logging.getLogger(__name__)
+
+# Private router, protected by API Key authentication
+router = APIRouter()
+
+
+async def is_websocket_alive(ws: WebSocket) -> bool:
+    """
+    Check if a FastAPI Websocket is still open
+    """
+    if (
+        ws.application_state == WebSocketState.CONNECTED and
+        ws.client_state == WebSocketState.CONNECTED
+    ):
+        return True
+    return False
+
+
+async def _process_consumer_request(
+    request: Dict[str, Any],
+    channel: WebSocketChannel,
+    rpc: RPC
+):
+    """
+    Validate and handle a request from a websocket consumer
+    """
+    # Validate the request, makes sure it matches the schema
+    try:
+        websocket_request = WSRequestSchema.parse_obj(request)
+    except ValidationError as e:
+        logger.error(f"Invalid request from {channel}: {e}")
+        return
+
+    type, data = websocket_request.type, websocket_request.data
+    response: WSMessageSchema
+
+    logger.debug(f"Request of type {type} from {channel}")
+
+    # If we have a request of type SUBSCRIBE, set the topics in this channel
+    if type == RPCRequestType.SUBSCRIBE:
+        # If the request is empty, do nothing
+        if not data:
+            return
+
+        # If all topics passed are a valid RPCMessageType, set subscriptions on channel
+        if all([any(x.value == topic for x in RPCMessageType) for topic in data]):
+            channel.set_subscriptions(data)
+
+        # We don't send a response for subscriptions
+        return
+
+    elif type == RPCRequestType.WHITELIST:
+        # Get whitelist
+        whitelist = rpc._ws_request_whitelist()
+
+        # Format response
+        response = WSWhitelistMessage(data=whitelist)
+        # Send it back
+        await channel.send(response.dict(exclude_none=True))
+
+    elif type == RPCRequestType.ANALYZED_DF:
+        limit = None
+
+        if data:
+            # Limit the amount of candles per dataframe to 'limit' or 1500
+            limit = max(data.get('limit', 1500), 1500)
+
+        # They requested the full historical analyzed dataframes
+        analyzed_df = rpc._ws_request_analyzed_df(limit)
+
+        # For every dataframe, send as a separate message
+        for _, message in analyzed_df.items():
+            response = WSAnalyzedDFMessage(data=message)
+            await channel.send(response.dict(exclude_none=True))
+
+
+@router.websocket("/message/ws")
+async def message_endpoint(
+    ws: WebSocket,
+    rpc: RPC = Depends(get_rpc),
+    channel_manager=Depends(get_channel_manager),
+    token: str = Depends(validate_ws_token)
+):
+    """
+    Message WebSocket endpoint, facilitates sending RPC messages
+    """
+    try:
+        channel = await channel_manager.on_connect(ws)
+
+        if await is_websocket_alive(ws):
+
+            logger.info(f"Consumer connected - {channel}")
+
+            # Keep connection open until explicitly closed, and process requests
+            try:
+                while not channel.is_closed():
+                    request = await channel.recv()
+
+                    # Process the request here
+                    await _process_consumer_request(request, channel, rpc)
+
+            except WebSocketDisconnect:
+                # Handle client disconnects
+                logger.info(f"Consumer disconnected - {channel}")
+                await channel_manager.on_disconnect(ws)
+            except Exception as e:
+                logger.info(f"Consumer connection failed - {channel}")
+                logger.exception(e)
+                # Handle cases like -
+                # RuntimeError('Cannot call "send" once a closed message has been sent')
+                await channel_manager.on_disconnect(ws)
+
+        else:
+            await ws.close()
+
+    except RuntimeError:
+        # WebSocket was closed
+        await channel_manager.on_disconnect(ws)
+
+    except Exception as e:
+        logger.error(f"Failed to serve - {ws.client}")
+        # Log tracebacks to keep track of what errors are happening
+        logger.exception(e)
+        await channel_manager.on_disconnect(ws)

freqtrade/rpc/api_server/deps.py

@@ -41,6 +41,10 @@ def get_exchange(config=Depends(get_config)):
     return ApiServer._exchange
 
 
+def get_channel_manager():
+    return ApiServer._ws_channel_manager
+
+
 def is_webserver_mode(config=Depends(get_config)):
     if config['runmode'] != RunMode.WEBSERVER:
         raise RPCException('Bot is not in the correct state')

freqtrade/rpc/api_server/webserver.py

@@ -1,16 +1,21 @@
+import asyncio
 import logging
 from ipaddress import IPv4Address
+from threading import Thread
 from typing import Any, Dict
 
 import orjson
 import uvicorn
 from fastapi import Depends, FastAPI
 from fastapi.middleware.cors import CORSMiddleware
+# Look into alternatives
+from janus import Queue as ThreadedQueue
 from starlette.responses import JSONResponse
 
 from freqtrade.constants import Config
 from freqtrade.exceptions import OperationalException
 from freqtrade.rpc.api_server.uvicorn_threaded import UvicornServer
+from freqtrade.rpc.api_server.ws import ChannelManager
 from freqtrade.rpc.rpc import RPC, RPCException, RPCHandler
 
 
@@ -44,6 +49,10 @@ class ApiServer(RPCHandler):
     _config: Config = {}
     # Exchange - only available in webserver mode.
     _exchange = None
+    # websocket message queue stuff
+    _ws_channel_manager = None
+    _ws_thread = None
+    _ws_loop = None
 
     def __new__(cls, *args, **kwargs):
         """
@@ -61,17 +70,21 @@ class ApiServer(RPCHandler):
             return
         self._standalone: bool = standalone
         self._server = None
+        self._ws_queue = None
+        self._ws_background_task = None
+
         ApiServer.__initialized = True
 
         api_config = self._config['api_server']
 
+        ApiServer._ws_channel_manager = ChannelManager()
+
         self.app = FastAPI(title="Freqtrade API",
                            docs_url='/docs' if api_config.get('enable_openapi', False) else None,
                            redoc_url=None,
                            default_response_class=FTJSONResponse,
                            )
         self.configure_app(self.app, self._config)
-
         self.start_api()
 
     def add_rpc_handler(self, rpc: RPC):
@@ -93,6 +106,19 @@ class ApiServer(RPCHandler):
             logger.info("Stopping API Server")
             self._server.cleanup()
 
+        if self._ws_thread and self._ws_loop:
+            logger.info("Stopping API Server background tasks")
+
+            if self._ws_background_task:
+                # Cancel the queue task
+                self._ws_background_task.cancel()
+
+            self._ws_thread.join()
+
+        self._ws_thread = None
+        self._ws_loop = None
+        self._ws_background_task = None
+
     @classmethod
     def shutdown(cls):
         cls.__initialized = False
@@ -102,7 +128,9 @@ class ApiServer(RPCHandler):
         cls._rpc = None
 
     def send_msg(self, msg: Dict[str, str]) -> None:
-        pass
+        if self._ws_queue:
+            sync_q = self._ws_queue.sync_q
+            sync_q.put(msg)
 
     def handle_rpc_exception(self, request, exc):
         logger.exception(f"API Error calling: {exc}")
@@ -116,6 +144,7 @@ class ApiServer(RPCHandler):
         from freqtrade.rpc.api_server.api_backtest import router as api_backtest
         from freqtrade.rpc.api_server.api_v1 import router as api_v1
         from freqtrade.rpc.api_server.api_v1 import router_public as api_v1_public
+        from freqtrade.rpc.api_server.api_ws import router as ws_router
         from freqtrade.rpc.api_server.web_ui import router_ui
 
         app.include_router(api_v1_public, prefix="/api/v1")
@@ -126,6 +155,7 @@ class ApiServer(RPCHandler):
         app.include_router(api_backtest, prefix="/api/v1",
                            dependencies=[Depends(http_basic_or_jwt_token)],
                            )
+        app.include_router(ws_router, prefix="/api/v1")
         app.include_router(router_login, prefix="/api/v1", tags=["auth"])
         # UI Router MUST be last!
         app.include_router(router_ui, prefix='')
@@ -140,6 +170,48 @@ class ApiServer(RPCHandler):
 
         app.add_exception_handler(RPCException, self.handle_rpc_exception)
 
+    def start_message_queue(self):
+        if self._ws_thread:
+            return
+
+        # Create a new loop, as it'll be just for the background thread
+        self._ws_loop = asyncio.new_event_loop()
+
+        # Start the thread
+        self._ws_thread = Thread(target=self._ws_loop.run_forever)
+        self._ws_thread.start()
+
+        # Finally, submit the coro to the thread
+        self._ws_background_task = asyncio.run_coroutine_threadsafe(
+            self._broadcast_queue_data(), loop=self._ws_loop)
+
+    async def _broadcast_queue_data(self):
+        # Instantiate the queue in this coroutine so it's attached to our loop
+        self._ws_queue = ThreadedQueue()
+        async_queue = self._ws_queue.async_q
+
+        try:
+            while True:
+                logger.debug("Getting queue messages...")
+                # Get data from queue
+                message = await async_queue.get()
+                logger.debug(f"Found message of type: {message.get('type')}")
+                # Broadcast it
+                await self._ws_channel_manager.broadcast(message)
+                # Sleep, make this configurable?
+                await asyncio.sleep(0.1)
+        except asyncio.CancelledError:
+            pass
+
+        # For testing, shouldn't happen when stable
+        except Exception as e:
+            logger.exception(f"Exception happened in background task: {e}")
+
+        finally:
+            # Disconnect channels and stop the loop on cancel
+            await self._ws_channel_manager.disconnect_all()
+            self._ws_loop.stop()
+
     def start_api(self):
         """
         Start API ... should be run in thread.
@@ -177,6 +249,7 @@ class ApiServer(RPCHandler):
             if self._standalone:
                 self._server.run()
             else:
+                self.start_message_queue()
                 self._server.run_in_thread()
         except Exception:
             logger.exception("Api server failed to start.")

freqtrade/rpc/api_server/ws/__init__.py

@@ -0,0 +1,6 @@
+# flake8: noqa: F401
+# isort: off
+from freqtrade.rpc.api_server.ws.types import WebSocketType
+from freqtrade.rpc.api_server.ws.proxy import WebSocketProxy
+from freqtrade.rpc.api_server.ws.serializer import HybridJSONWebSocketSerializer
+from freqtrade.rpc.api_server.ws.channel import ChannelManager, WebSocketChannel

freqtrade/rpc/api_server/ws/channel.py

@@ -0,0 +1,178 @@
+import logging
+from threading import RLock
+from typing import List, Optional, Type
+from uuid import uuid4
+
+from fastapi import WebSocket as FastAPIWebSocket
+
+from freqtrade.rpc.api_server.ws.proxy import WebSocketProxy
+from freqtrade.rpc.api_server.ws.serializer import (HybridJSONWebSocketSerializer,
+                                                    WebSocketSerializer)
+from freqtrade.rpc.api_server.ws.types import WebSocketType
+
+
+logger = logging.getLogger(__name__)
+
+
+class WebSocketChannel:
+    """
+    Object to help facilitate managing a websocket connection
+    """
+
+    def __init__(
+        self,
+        websocket: WebSocketType,
+        channel_id: Optional[str] = None,
+        serializer_cls: Type[WebSocketSerializer] = HybridJSONWebSocketSerializer
+    ):
+
+        self.channel_id = channel_id if channel_id else uuid4().hex[:8]
+
+        # The WebSocket object
+        self._websocket = WebSocketProxy(websocket)
+        # The Serializing class for the WebSocket object
+        self._serializer_cls = serializer_cls
+
+        self._subscriptions: List[str] = []
+
+        # Internal event to signify a closed websocket
+        self._closed = False
+
+        # Wrap the WebSocket in the Serializing class
+        self._wrapped_ws = self._serializer_cls(self._websocket)
+
+    def __repr__(self):
+        return f"WebSocketChannel({self.channel_id}, {self.remote_addr})"
+
+    @property
+    def remote_addr(self):
+        return self._websocket.remote_addr
+
+    async def send(self, data):
+        """
+        Send data on the wrapped websocket
+        """
+        await self._wrapped_ws.send(data)
+
+    async def recv(self):
+        """
+        Receive data on the wrapped websocket
+        """
+        return await self._wrapped_ws.recv()
+
+    async def ping(self):
+        """
+        Ping the websocket
+        """
+        return await self._websocket.ping()
+
+    async def close(self):
+        """
+        Close the WebSocketChannel
+        """
+
+        self._closed = True
+
+    def is_closed(self) -> bool:
+        """
+        Closed flag
+        """
+        return self._closed
+
+    def set_subscriptions(self, subscriptions: List[str] = []) -> None:
+        """
+        Set which subscriptions this channel is subscribed to
+
+        :param subscriptions: List of subscriptions, List[str]
+        """
+        self._subscriptions = subscriptions
+
+    def subscribed_to(self, message_type: str) -> bool:
+        """
+        Check if this channel is subscribed to the message_type
+
+        :param message_type: The message type to check
+        """
+        return message_type in self._subscriptions
+
+
+class ChannelManager:
+    def __init__(self):
+        self.channels = dict()
+        self._lock = RLock()  # Re-entrant Lock
+
+    async def on_connect(self, websocket: WebSocketType):
+        """
+        Wrap websocket connection into Channel and add to list
+
+        :param websocket: The WebSocket object to attach to the Channel
+        """
+        if isinstance(websocket, FastAPIWebSocket):
+            try:
+                await websocket.accept()
+            except RuntimeError:
+                # The connection was closed before we could accept it
+                return
+
+        ws_channel = WebSocketChannel(websocket)
+
+        with self._lock:
+            self.channels[websocket] = ws_channel
+
+        return ws_channel
+
+    async def on_disconnect(self, websocket: WebSocketType):
+        """
+        Call close on the channel if it's not, and remove from channel list
+
+        :param websocket: The WebSocket objet attached to the Channel
+        """
+        with self._lock:
+            channel = self.channels.get(websocket)
+            if channel:
+                if not channel.is_closed():
+                    await channel.close()
+
+                del self.channels[websocket]
+
+    async def disconnect_all(self):
+        """
+        Disconnect all Channels
+        """
+        with self._lock:
+            for websocket, channel in self.channels.items():
+                if not channel.is_closed():
+                    await channel.close()
+
+            self.channels = dict()
+
+    async def broadcast(self, data):
+        """
+        Broadcast data on all Channels
+
+        :param data: The data to send
+        """
+        with self._lock:
+            message_type = data.get('type')
+            for websocket, channel in self.channels.items():
+                try:
+                    if channel.subscribed_to(message_type):
+                        await channel.send(data)
+                except RuntimeError:
+                    # Handle cannot send after close cases
+                    await self.on_disconnect(websocket)
+
+    async def send_direct(self, channel, data):
+        """
+        Send data directly through direct_channel only
+
+        :param direct_channel: The WebSocketChannel object to send data through
+        :param data: The data to send
+        """
+        await channel.send(data)
+
+    def has_channels(self):
+        """
+        Flag for more than 0 channels
+        """
+        return len(self.channels) > 0

freqtrade/rpc/api_server/ws/proxy.py

@@ -0,0 +1,69 @@
+from typing import Any, Tuple, Union
+
+from fastapi import WebSocket as FastAPIWebSocket
+from websockets.client import WebSocketClientProtocol as WebSocket
+
+from freqtrade.rpc.api_server.ws.types import WebSocketType
+
+
+class WebSocketProxy:
+    """
+    WebSocketProxy object to bring the FastAPIWebSocket and websockets.WebSocketClientProtocol
+    under the same API
+    """
+
+    def __init__(self, websocket: WebSocketType):
+        self._websocket: Union[FastAPIWebSocket, WebSocket] = websocket
+
+    @property
+    def remote_addr(self) -> Tuple[Any, ...]:
+        if isinstance(self._websocket, WebSocket):
+            return self._websocket.remote_address
+        elif isinstance(self._websocket, FastAPIWebSocket):
+            if self._websocket.client:
+                client, port = self._websocket.client.host, self._websocket.client.port
+                return (client, port)
+        return ("unknown", 0)
+
+    async def send(self, data):
+        """
+        Send data on the wrapped websocket
+        """
+        if hasattr(self._websocket, "send_text"):
+            await self._websocket.send_text(data)
+        else:
+            await self._websocket.send(data)
+
+    async def recv(self):
+        """
+        Receive data on the wrapped websocket
+        """
+        if hasattr(self._websocket, "receive_text"):
+            return await self._websocket.receive_text()
+        else:
+            return await self._websocket.recv()
+
+    async def ping(self):
+        """
+        Ping the websocket, not supported by FastAPI WebSockets
+        """
+        if hasattr(self._websocket, "ping"):
+            return await self._websocket.ping()
+        return False
+
+    async def close(self, code: int = 1000):
+        """
+        Close the websocket connection, only supported by FastAPI WebSockets
+        """
+        if hasattr(self._websocket, "close"):
+            try:
+                return await self._websocket.close(code)
+            except RuntimeError:
+                pass
+
+    async def accept(self):
+        """
+        Accept the WebSocket connection, only support by FastAPI WebSockets
+        """
+        if hasattr(self._websocket, "accept"):
+            return await self._websocket.accept()

freqtrade/rpc/api_server/ws/serializer.py

@@ -0,0 +1,62 @@
+import logging
+from abc import ABC, abstractmethod
+
+import orjson
+import rapidjson
+from pandas import DataFrame
+
+from freqtrade.misc import dataframe_to_json, json_to_dataframe
+from freqtrade.rpc.api_server.ws.proxy import WebSocketProxy
+
+
+logger = logging.getLogger(__name__)
+
+
+class WebSocketSerializer(ABC):
+    def __init__(self, websocket: WebSocketProxy):
+        self._websocket: WebSocketProxy = websocket
+
+    @abstractmethod
+    def _serialize(self, data):
+        raise NotImplementedError()
+
+    @abstractmethod
+    def _deserialize(self, data):
+        raise NotImplementedError()
+
+    async def send(self, data: bytes):
+        await self._websocket.send(self._serialize(data))
+
+    async def recv(self) -> bytes:
+        data = await self._websocket.recv()
+
+        return self._deserialize(data)
+
+    async def close(self, code: int = 1000):
+        await self._websocket.close(code)
+
+
+class HybridJSONWebSocketSerializer(WebSocketSerializer):
+    def _serialize(self, data) -> str:
+        return str(orjson.dumps(data, default=_json_default), "utf-8")
+
+    def _deserialize(self, data: str):
+        # RapidJSON expects strings
+        return rapidjson.loads(data, object_hook=_json_object_hook)
+
+
+# Support serializing pandas DataFrames
+def _json_default(z):
+    if isinstance(z, DataFrame):
+        return {
+            '__type__': 'dataframe',
+            '__value__': dataframe_to_json(z)
+        }
+    raise TypeError
+
+
+# Support deserializing JSON to pandas DataFrames
+def _json_object_hook(z):
+    if z.get('__type__') == 'dataframe':
+        return json_to_dataframe(z.get('__value__'))
+    return z

freqtrade/rpc/api_server/ws/types.py

@@ -0,0 +1,8 @@
+from typing import Any, Dict, TypeVar
+
+from fastapi import WebSocket as FastAPIWebSocket
+from websockets.client import WebSocketClientProtocol as WebSocket
+
+
+WebSocketType = TypeVar("WebSocketType", FastAPIWebSocket, WebSocket)
+MessageType = Dict[str, Any]

freqtrade/rpc/api_server/ws_schemas.py

@@ -0,0 +1,63 @@
+from datetime import datetime
+from typing import Any, Dict, List, Optional
+
+from pandas import DataFrame
+from pydantic import BaseModel
+
+from freqtrade.constants import PairWithTimeframe
+from freqtrade.enums.rpcmessagetype import RPCMessageType, RPCRequestType
+
+
+class BaseArbitraryModel(BaseModel):
+    class Config:
+        arbitrary_types_allowed = True
+
+
+class WSRequestSchema(BaseArbitraryModel):
+    type: RPCRequestType
+    data: Optional[Any] = None
+
+
+class WSMessageSchema(BaseArbitraryModel):
+    type: RPCMessageType
+    data: Optional[Any] = None
+
+    class Config:
+        extra = 'allow'
+
+
+# ------------------------------ REQUEST SCHEMAS ----------------------------
+
+
+class WSSubscribeRequest(WSRequestSchema):
+    type: RPCRequestType = RPCRequestType.SUBSCRIBE
+    data: List[RPCMessageType]
+
+
+class WSWhitelistRequest(WSRequestSchema):
+    type: RPCRequestType = RPCRequestType.WHITELIST
+    data: None = None
+
+
+class WSAnalyzedDFRequest(WSRequestSchema):
+    type: RPCRequestType = RPCRequestType.ANALYZED_DF
+    data: Dict[str, Any] = {"limit": 1500}
+
+
+# ------------------------------ MESSAGE SCHEMAS ----------------------------
+
+class WSWhitelistMessage(WSMessageSchema):
+    type: RPCMessageType = RPCMessageType.WHITELIST
+    data: List[str]
+
+
+class WSAnalyzedDFMessage(WSMessageSchema):
+    class AnalyzedDFData(BaseArbitraryModel):
+        key: PairWithTimeframe
+        df: DataFrame
+        la: datetime
+
+    type: RPCMessageType = RPCMessageType.ANALYZED_DF
+    data: AnalyzedDFData
+
+# --------------------------------------------------------------------------

freqtrade/rpc/external_message_consumer.py

@@ -0,0 +1,335 @@
+"""
+ExternalMessageConsumer module
+
+Main purpose is to connect to external bot's message websocket to consume data
+from it
+"""
+import asyncio
+import logging
+import socket
+from threading import Thread
+from typing import TYPE_CHECKING, Any, Callable, Dict, List, TypedDict
+
+import websockets
+from pydantic import ValidationError
+
+from freqtrade.data.dataprovider import DataProvider
+from freqtrade.enums import RPCMessageType
+from freqtrade.misc import remove_entry_exit_signals
+from freqtrade.rpc.api_server.ws import WebSocketChannel
+from freqtrade.rpc.api_server.ws_schemas import (WSAnalyzedDFMessage, WSAnalyzedDFRequest,
+                                                 WSMessageSchema, WSRequestSchema,
+                                                 WSSubscribeRequest, WSWhitelistMessage,
+                                                 WSWhitelistRequest)
+
+
+if TYPE_CHECKING:
+    import websockets.connect
+
+
+class Producer(TypedDict):
+    name: str
+    host: str
+    port: int
+    ws_token: str
+
+
+logger = logging.getLogger(__name__)
+
+
+class ExternalMessageConsumer:
+    """
+    The main controller class for consuming external messages from
+    other freqtrade bot's
+    """
+
+    def __init__(
+        self,
+        config: Dict[str, Any],
+        dataprovider: DataProvider
+    ):
+        self._config = config
+        self._dp = dataprovider
+
+        self._running = False
+        self._thread = None
+        self._loop = None
+        self._main_task = None
+        self._sub_tasks = None
+
+        self._emc_config = self._config.get('external_message_consumer', {})
+
+        self.enabled = self._emc_config.get('enabled', False)
+        self.producers: List[Producer] = self._emc_config.get('producers', [])
+
+        self.wait_timeout = self._emc_config.get('wait_timeout', 300)  # in seconds
+        self.ping_timeout = self._emc_config.get('ping_timeout', 10)  # in seconds
+        self.sleep_time = self._emc_config.get('sleep_time', 10)  # in seconds
+
+        # The amount of candles per dataframe on the initial request
+        self.initial_candle_limit = self._emc_config.get('initial_candle_limit', 1500)
+
+        # Message size limit, in megabytes. Default 8mb, Use bitwise operator << 20 to convert
+        # as the websockets client expects bytes.
+        self.message_size_limit = (self._emc_config.get('message_size_limit', 8) << 20)
+
+        # Setting these explicitly as they probably shouldn't be changed by a user
+        # Unless we somehow integrate this with the strategy to allow creating
+        # callbacks for the messages
+        self.topics = [RPCMessageType.WHITELIST, RPCMessageType.ANALYZED_DF]
+
+        # Allow setting data for each initial request
+        self._initial_requests: List[WSRequestSchema] = [
+            WSSubscribeRequest(data=self.topics),
+            WSWhitelistRequest(),
+            WSAnalyzedDFRequest()
+        ]
+
+        # Specify which function to use for which RPCMessageType
+        self._message_handlers: Dict[str, Callable[[str, WSMessageSchema], None]] = {
+            RPCMessageType.WHITELIST: self._consume_whitelist_message,
+            RPCMessageType.ANALYZED_DF: self._consume_analyzed_df_message,
+        }
+
+        self.start()
+
+    def start(self):
+        """
+        Start the main internal loop in another thread to run coroutines
+        """
+        if self._thread and self._loop:
+            return
+
+        logger.info("Starting ExternalMessageConsumer")
+
+        self._loop = asyncio.new_event_loop()
+        self._thread = Thread(target=self._loop.run_forever)
+        self._running = True
+        self._thread.start()
+
+        self._main_task = asyncio.run_coroutine_threadsafe(self._main(), loop=self._loop)
+
+    def shutdown(self):
+        """
+        Shutdown the loop, thread, and tasks
+        """
+        if self._thread and self._loop:
+            logger.info("Stopping ExternalMessageConsumer")
+            self._running = False
+
+            if self._sub_tasks:
+                # Cancel sub tasks
+                for task in self._sub_tasks:
+                    task.cancel()
+
+            if self._main_task:
+                # Cancel the main task
+                self._main_task.cancel()
+
+            self._thread.join()
+
+            self._thread = None
+            self._loop = None
+            self._sub_tasks = None
+            self._main_task = None
+
+    async def _main(self):
+        """
+        The main task coroutine
+        """
+        lock = asyncio.Lock()
+
+        try:
+            # Create a connection to each producer
+            self._sub_tasks = [
+                self._loop.create_task(self._handle_producer_connection(producer, lock))
+                for producer in self.producers
+            ]
+
+            await asyncio.gather(*self._sub_tasks)
+        except asyncio.CancelledError:
+            pass
+        finally:
+            # Stop the loop once we are done
+            self._loop.stop()
+
+    async def _handle_producer_connection(self, producer: Producer, lock: asyncio.Lock):
+        """
+        Main connection loop for the consumer
+
+        :param producer: Dictionary containing producer info
+        :param lock: An asyncio Lock
+        """
+        try:
+            await self._create_connection(producer, lock)
+        except asyncio.CancelledError:
+            # Exit silently
+            pass
+
+    async def _create_connection(self, producer: Producer, lock: asyncio.Lock):
+        """
+        Actually creates and handles the websocket connection, pinging on timeout
+        and handling connection errors.
+
+        :param producer: Dictionary containing producer info
+        :param lock: An asyncio Lock
+        """
+        while self._running:
+            try:
+                host, port = producer['host'], producer['port']
+                token = producer['ws_token']
+                name = producer['name']
+                ws_url = f"ws://{host}:{port}/api/v1/message/ws?token={token}"
+
+                # This will raise InvalidURI if the url is bad
+                async with websockets.connect(ws_url, max_size=self.message_size_limit) as ws:
+                    channel = WebSocketChannel(ws, channel_id=name)
+
+                    logger.info(f"Producer connection success - {channel}")
+
+                    # Now request the initial data from this Producer
+                    for request in self._initial_requests:
+                        await channel.send(
+                            request.dict(exclude_none=True)
+                        )
+
+                    # Now receive data, if none is within the time limit, ping
+                    await self._receive_messages(channel, producer, lock)
+
+            except (websockets.exceptions.InvalidURI, ValueError) as e:
+                logger.error(f"{ws_url} is an invalid WebSocket URL - {e}")
+                break
+
+            except (
+                socket.gaierror,
+                ConnectionRefusedError,
+                websockets.exceptions.InvalidStatusCode,
+                websockets.exceptions.InvalidMessage
+            ) as e:
+                logger.error(f"Connection Refused - {e} retrying in {self.sleep_time}s")
+                await asyncio.sleep(self.sleep_time)
+                continue
+
+            except (
+                websockets.exceptions.ConnectionClosedError,
+                websockets.exceptions.ConnectionClosedOK
+            ):
+                # Just keep trying to connect again indefinitely
+                await asyncio.sleep(self.sleep_time)
+                continue
+
+            except Exception as e:
+                # An unforseen error has occurred, log and continue
+                logger.error("Unexpected error has occurred:")
+                logger.exception(e)
+                continue
+
+    async def _receive_messages(
+        self,
+        channel: WebSocketChannel,
+        producer: Producer,
+        lock: asyncio.Lock
+    ):
+        """
+        Loop to handle receiving messages from a Producer
+
+        :param channel: The WebSocketChannel object for the WebSocket
+        :param producer: Dictionary containing producer info
+        :param lock: An asyncio Lock
+        """
+        while self._running:
+            try:
+                message = await asyncio.wait_for(
+                    channel.recv(),
+                    timeout=self.wait_timeout
+                )
+
+                try:
+                    async with lock:
+                        # Handle the message
+                        self.handle_producer_message(producer, message)
+                except Exception as e:
+                    logger.exception(f"Error handling producer message: {e}")
+
+            except (asyncio.TimeoutError, websockets.exceptions.ConnectionClosed):
+                # We haven't received data yet. Check the connection and continue.
+                try:
+                    # ping
+                    ping = await channel.ping()
+
+                    await asyncio.wait_for(ping, timeout=self.ping_timeout)
+                    logger.debug(f"Connection to {channel} still alive...")
+
+                    continue
+                except Exception as e:
+                    logger.warning(f"Ping error {channel} - retrying in {self.sleep_time}s")
+                    logger.debug(e, exc_info=e)
+                    await asyncio.sleep(self.sleep_time)
+
+                    break
+
+    def handle_producer_message(self, producer: Producer, message: Dict[str, Any]):
+        """
+        Handles external messages from a Producer
+        """
+        producer_name = producer.get('name', 'default')
+
+        try:
+            producer_message = WSMessageSchema.parse_obj(message)
+        except ValidationError as e:
+            logger.error(f"Invalid message from `{producer_name}`: {e}")
+            return
+
+        if not producer_message.data:
+            logger.error(f"Empty message received from `{producer_name}`")
+            return
+
+        logger.info(f"Received message of type `{producer_message.type}` from `{producer_name}`")
+
+        message_handler = self._message_handlers.get(producer_message.type)
+
+        if not message_handler:
+            logger.info(f"Received unhandled message: `{producer_message.data}`, ignoring...")
+            return
+
+        message_handler(producer_name, producer_message)
+
+    def _consume_whitelist_message(self, producer_name: str, message: WSMessageSchema):
+        try:
+            # Validate the message
+            whitelist_message = WSWhitelistMessage.parse_obj(message)
+        except ValidationError as e:
+            logger.error(f"Invalid message from `{producer_name}`: {e}")
+            return
+
+        # Add the pairlist data to the DataProvider
+        self._dp._set_producer_pairs(whitelist_message.data, producer_name=producer_name)
+
+        logger.debug(f"Consumed message from `{producer_name}` of type `RPCMessageType.WHITELIST`")
+
+    def _consume_analyzed_df_message(self, producer_name: str, message: WSMessageSchema):
+        try:
+            df_message = WSAnalyzedDFMessage.parse_obj(message)
+        except ValidationError as e:
+            logger.error(f"Invalid message from `{producer_name}`: {e}")
+            return
+
+        key = df_message.data.key
+        df = df_message.data.df
+        la = df_message.data.la
+
+        pair, timeframe, candle_type = key
+
+        # If set, remove the Entry and Exit signals from the Producer
+        if self._emc_config.get('remove_entry_exit_signals', False):
+            df = remove_entry_exit_signals(df)
+
+        # Add the dataframe to the dataprovider
+        self._dp._add_external_df(pair, df,
+                                  last_analyzed=la,
+                                  timeframe=timeframe,
+                                  candle_type=candle_type,
+                                  producer_name=producer_name)
+
+        logger.debug(
+            f"Consumed message from `{producer_name}` of type `RPCMessageType.ANALYZED_DF`")

freqtrade/rpc/rpc.py

@@ -1039,14 +1039,52 @@ class RPC:
 
     def _rpc_analysed_dataframe(self, pair: str, timeframe: str,
                                 limit: Optional[int]) -> Dict[str, Any]:
+        """ Analyzed dataframe in Dict form """
 
+        _data, last_analyzed = self.__rpc_analysed_dataframe_raw(pair, timeframe, limit)
+        return self._convert_dataframe_to_dict(self._freqtrade.config['strategy'],
+                                               pair, timeframe, _data, last_analyzed)
+
+    def __rpc_analysed_dataframe_raw(self, pair: str, timeframe: str,
+                                     limit: Optional[int]) -> Tuple[DataFrame, datetime]:
+        """ Get the dataframe and last analyze from the dataprovider """
         _data, last_analyzed = self._freqtrade.dataprovider.get_analyzed_dataframe(
             pair, timeframe)
         _data = _data.copy()
+
         if limit:
             _data = _data.iloc[-limit:]
-        return self._convert_dataframe_to_dict(self._freqtrade.config['strategy'],
+        return _data, last_analyzed
-                                               pair, timeframe, _data, last_analyzed)
+
+    def _ws_all_analysed_dataframes(
+        self,
+        pairlist: List[str],
+        limit: Optional[int]
+    ) -> Dict[str, Any]:
+        """ Get the analysed dataframes of each pair in the pairlist """
+        timeframe = self._freqtrade.config['timeframe']
+        candle_type = self._freqtrade.config.get('candle_type_def', CandleType.SPOT)
+        _data = {}
+
+        for pair in pairlist:
+            dataframe, last_analyzed = self.__rpc_analysed_dataframe_raw(pair, timeframe, limit)
+
+            _data[pair] = {
+                "key": (pair, timeframe, candle_type),
+                "df": dataframe,
+                "la": last_analyzed
+            }
+
+        return _data
+
+    def _ws_request_analyzed_df(self, limit: Optional[int]):
+        """ Historical Analyzed Dataframes for WebSocket """
+        whitelist = self._freqtrade.active_pair_whitelist
+        return self._ws_all_analysed_dataframes(whitelist, limit)
+
+    def _ws_request_whitelist(self):
+        """ Whitelist data for WebSocket """
+        return self._freqtrade.active_pair_whitelist
 
     @staticmethod
     def _rpc_analysed_history_full(config, pair: str, timeframe: str,

freqtrade/rpc/rpc_manager.py

@@ -67,7 +67,8 @@ class RPCManager:
             'status': 'stopping bot'
         }
         """
-        logger.info('Sending rpc message: %s', msg)
+        if msg.get('type') is not RPCMessageType.ANALYZED_DF:
+            logger.info('Sending rpc message: %s', msg)
         if 'pair' in msg:
             msg.update({
                 'base_currency': self._rpc._freqtrade.exchange.get_pair_base_currency(msg['pair'])

freqtrade/strategy/interface.py

@@ -16,6 +16,7 @@ from freqtrade.enums import (CandleType, ExitCheckTuple, ExitType, RunMode, Sign
                              SignalTagType, SignalType, TradingMode)
 from freqtrade.exceptions import OperationalException, StrategyError
 from freqtrade.exchange import timeframe_to_minutes, timeframe_to_next_date, timeframe_to_seconds
+from freqtrade.misc import remove_entry_exit_signals
 from freqtrade.persistence import Order, PairLocks, Trade
 from freqtrade.strategy.hyper import HyperStrategyMixin
 from freqtrade.strategy.informative_decorator import (InformativeData, PopulateIndicators,
@@ -742,20 +743,19 @@ class IStrategy(ABC, HyperStrategyMixin):
         # always run if process_only_new_candles is set to false
         if (not self.process_only_new_candles or
                 self._last_candle_seen_per_pair.get(pair, None) != dataframe.iloc[-1]['date']):
+
             # Defs that only make change on new candle data.
             dataframe = self.analyze_ticker(dataframe, metadata)
+
             self._last_candle_seen_per_pair[pair] = dataframe.iloc[-1]['date']
-            self.dp._set_cached_df(
+
-                pair, self.timeframe, dataframe,
+            candle_type = self.config.get('candle_type_def', CandleType.SPOT)
-                candle_type=self.config.get('candle_type_def', CandleType.SPOT))
+            self.dp._set_cached_df(pair, self.timeframe, dataframe, candle_type=candle_type)
+            self.dp._emit_df((pair, self.timeframe, candle_type), dataframe)
+
         else:
             logger.debug("Skipping TA Analysis for already analyzed candle")
-            dataframe[SignalType.ENTER_LONG.value] = 0
+            dataframe = remove_entry_exit_signals(dataframe)
-            dataframe[SignalType.EXIT_LONG.value] = 0
-            dataframe[SignalType.ENTER_SHORT.value] = 0
-            dataframe[SignalType.EXIT_SHORT.value] = 0
-            dataframe[SignalTagType.ENTER_TAG.value] = None
-            dataframe[SignalTagType.EXIT_TAG.value] = None
 
         logger.debug("Loop Analysis Launched")
 

freqtrade/templates/base_config.json.j2

@@ -67,6 +67,7 @@
         "verbosity": "error",
         "enable_openapi": false,
         "jwt_secret_key": "{{ api_server_jwt_key }}",
+        "ws_token": "{{ api_server_ws_token }}",
         "CORS_origins": [],
         "username": "{{ api_server_username }}",
         "password": "{{ api_server_password }}"

mkdocs.yml

@@ -28,9 +28,7 @@ nav:
         - Configuration: freqai-configuration.md
         - Parameter table: freqai-parameter-table.md
         - Feature engineering: freqai-feature-engineering.md
-        # - Outlier detection: freqai-outlier-detection.md
         - Running FreqAI: freqai-running.md
-        # - Data handling: freqai-data-handling.md
         - Developer guide: freqai-developers.md
     - Short / Leverage: leverage.md
     - Utility Sub-commands: utils.md
@@ -44,6 +42,7 @@ nav:
         - Advanced Post-installation Tasks: advanced-setup.md
         - Advanced Strategy: strategy-advanced.md
         - Advanced Hyperopt: advanced-hyperopt.md
+        - Producer/Consumer mode: producer-consumer.md
         - Edge Positioning: edge.md
         - Sandbox Testing: sandbox-testing.md
     - FAQ: faq.md

requirements.txt

@@ -21,6 +21,7 @@ jinja2==3.1.2
 tables==3.7.0
 blosc==1.10.6
 joblib==1.2.0
+pyarrow==9.0.0
 
 # find first, C search in arrays
 py_find_1st==1.1.5
@@ -50,3 +51,7 @@ python-dateutil==2.8.2
 
 #Futures
 schedule==1.1.0
+
+#WS Messages
+websockets==10.3
+janus==1.0.0

setup.cfg

@@ -49,4 +49,3 @@ exclude =
     __pycache__,
     .eggs,
     user_data,
-

setup.py

@@ -8,13 +8,11 @@ hyperopt = [
     'scikit-learn',
     'scikit-optimize>=0.7.0',
     'filelock',
-    'joblib',
     'progressbar2',
 ]
 
 freqai = [
     'scikit-learn',
-    'joblib',
     'catboost; platform_machine != "aarch64"',
     'lightgbm',
 ]
@@ -74,12 +72,16 @@ setup(
         'pandas',
         'tables',
         'blosc',
+        'joblib',
+        'pyarrow',
         'fastapi',
         'uvicorn',
         'psutil',
         'pyjwt',
         'aiofiles',
-        'schedule'
+        'schedule',
+        'websockets',
+        'janus'
     ],
     extras_require={
         'dev': all_extra,

tests/conftest.py

@@ -58,6 +58,11 @@ def log_has(line, logs):
     return any(line == message for message in logs.messages)
 
 
+def log_has_when(line, logs, when):
+    """Check if line is found in caplog's messages during a specified stage"""
+    return any(line == message.message for message in logs.get_records(when))
+
+
 def log_has_re(line, logs):
     """Check if line matches some caplog's message."""
     return any(re.match(line, message) for message in logs.messages)

tests/data/test_datahandler.py

@@ -0,0 +1,436 @@
+# pragma pylint: disable=missing-docstring, protected-access, C0103
+
+import re
+from pathlib import Path
+from unittest.mock import MagicMock
+
+import pytest
+from pandas import DataFrame
+
+from freqtrade.configuration import TimeRange
+from freqtrade.constants import AVAILABLE_DATAHANDLERS
+from freqtrade.data.history.featherdatahandler import FeatherDataHandler
+from freqtrade.data.history.hdf5datahandler import HDF5DataHandler
+from freqtrade.data.history.idatahandler import IDataHandler, get_datahandler, get_datahandlerclass
+from freqtrade.data.history.jsondatahandler import JsonDataHandler, JsonGzDataHandler
+from freqtrade.data.history.parquetdatahandler import ParquetDataHandler
+from freqtrade.enums import CandleType, TradingMode
+from tests.conftest import log_has
+
+
+def test_datahandler_ohlcv_get_pairs(testdatadir):
+    pairs = JsonDataHandler.ohlcv_get_pairs(testdatadir, '5m', candle_type=CandleType.SPOT)
+    # Convert to set to avoid failures due to sorting
+    assert set(pairs) == {'UNITTEST/BTC', 'XLM/BTC', 'ETH/BTC', 'TRX/BTC', 'LTC/BTC',
+                          'XMR/BTC', 'ZEC/BTC', 'ADA/BTC', 'ETC/BTC', 'NXT/BTC',
+                          'DASH/BTC', 'XRP/ETH'}
+
+    pairs = JsonGzDataHandler.ohlcv_get_pairs(testdatadir, '8m', candle_type=CandleType.SPOT)
+    assert set(pairs) == {'UNITTEST/BTC'}
+
+    pairs = HDF5DataHandler.ohlcv_get_pairs(testdatadir, '5m', candle_type=CandleType.SPOT)
+    assert set(pairs) == {'UNITTEST/BTC'}
+
+    pairs = JsonDataHandler.ohlcv_get_pairs(testdatadir, '1h', candle_type=CandleType.MARK)
+    assert set(pairs) == {'UNITTEST/USDT', 'XRP/USDT'}
+
+    pairs = JsonGzDataHandler.ohlcv_get_pairs(testdatadir, '1h', candle_type=CandleType.FUTURES)
+    assert set(pairs) == {'XRP/USDT'}
+
+    pairs = HDF5DataHandler.ohlcv_get_pairs(testdatadir, '1h', candle_type=CandleType.MARK)
+    assert set(pairs) == {'UNITTEST/USDT:USDT'}
+
+
+@pytest.mark.parametrize('filename,pair,timeframe,candletype', [
+    ('XMR_BTC-5m.json', 'XMR_BTC', '5m', ''),
+    ('XMR_USDT-1h.h5', 'XMR_USDT', '1h', ''),
+    ('BTC-PERP-1h.h5', 'BTC-PERP', '1h', ''),
+    ('BTC_USDT-2h.jsongz', 'BTC_USDT', '2h', ''),
+    ('BTC_USDT-2h-mark.jsongz', 'BTC_USDT', '2h', 'mark'),
+    ('XMR_USDT-1h-mark.h5', 'XMR_USDT', '1h', 'mark'),
+    ('XMR_USDT-1h-random.h5', 'XMR_USDT', '1h', 'random'),
+    ('BTC-PERP-1h-index.h5', 'BTC-PERP', '1h', 'index'),
+    ('XMR_USDT_USDT-1h-mark.h5', 'XMR_USDT_USDT', '1h', 'mark'),
+])
+def test_datahandler_ohlcv_regex(filename, pair, timeframe, candletype):
+    regex = JsonDataHandler._OHLCV_REGEX
+
+    match = re.search(regex, filename)
+    assert len(match.groups()) > 1
+    assert match[1] == pair
+    assert match[2] == timeframe
+    assert match[3] == candletype
+
+
+@pytest.mark.parametrize('input,expected', [
+    ('XMR_USDT', 'XMR/USDT'),
+    ('BTC_USDT', 'BTC/USDT'),
+    ('USDT_BUSD', 'USDT/BUSD'),
+    ('BTC_USDT_USDT', 'BTC/USDT:USDT'),  # Futures
+    ('XRP_USDT_USDT', 'XRP/USDT:USDT'),  # futures
+    ('BTC-PERP', 'BTC-PERP'),
+    ('BTC-PERP_USDT', 'BTC-PERP:USDT'),  # potential FTX case
+    ('UNITTEST_USDT', 'UNITTEST/USDT'),
+])
+def test_rebuild_pair_from_filename(input, expected):
+
+    assert IDataHandler.rebuild_pair_from_filename(input) == expected
+
+
+def test_datahandler_ohlcv_get_available_data(testdatadir):
+    paircombs = JsonDataHandler.ohlcv_get_available_data(testdatadir, TradingMode.SPOT)
+    # Convert to set to avoid failures due to sorting
+    assert set(paircombs) == {
+        ('UNITTEST/BTC', '5m', CandleType.SPOT),
+        ('ETH/BTC', '5m', CandleType.SPOT),
+        ('XLM/BTC', '5m', CandleType.SPOT),
+        ('TRX/BTC', '5m', CandleType.SPOT),
+        ('LTC/BTC', '5m', CandleType.SPOT),
+        ('XMR/BTC', '5m', CandleType.SPOT),
+        ('ZEC/BTC', '5m', CandleType.SPOT),
+        ('UNITTEST/BTC', '1m', CandleType.SPOT),
+        ('ADA/BTC', '5m', CandleType.SPOT),
+        ('ETC/BTC', '5m', CandleType.SPOT),
+        ('NXT/BTC', '5m', CandleType.SPOT),
+        ('DASH/BTC', '5m', CandleType.SPOT),
+        ('XRP/ETH', '1m', CandleType.SPOT),
+        ('XRP/ETH', '5m', CandleType.SPOT),
+        ('UNITTEST/BTC', '30m', CandleType.SPOT),
+        ('UNITTEST/BTC', '8m', CandleType.SPOT),
+        ('NOPAIR/XXX', '4m', CandleType.SPOT),
+    }
+
+    paircombs = JsonDataHandler.ohlcv_get_available_data(testdatadir, TradingMode.FUTURES)
+    # Convert to set to avoid failures due to sorting
+    assert set(paircombs) == {
+        ('UNITTEST/USDT', '1h', 'mark'),
+        ('XRP/USDT', '1h', 'futures'),
+        ('XRP/USDT', '1h', 'mark'),
+        ('XRP/USDT', '8h', 'mark'),
+        ('XRP/USDT', '8h', 'funding_rate'),
+    }
+
+    paircombs = JsonGzDataHandler.ohlcv_get_available_data(testdatadir, TradingMode.SPOT)
+    assert set(paircombs) == {('UNITTEST/BTC', '8m', CandleType.SPOT)}
+    paircombs = HDF5DataHandler.ohlcv_get_available_data(testdatadir, TradingMode.SPOT)
+    assert set(paircombs) == {('UNITTEST/BTC', '5m', CandleType.SPOT)}
+
+
+def test_jsondatahandler_trades_get_pairs(testdatadir):
+    pairs = JsonGzDataHandler.trades_get_pairs(testdatadir)
+    # Convert to set to avoid failures due to sorting
+    assert set(pairs) == {'XRP/ETH', 'XRP/OLD'}
+
+
+def test_jsondatahandler_ohlcv_purge(mocker, testdatadir):
+    mocker.patch.object(Path, "exists", MagicMock(return_value=False))
+    unlinkmock = mocker.patch.object(Path, "unlink", MagicMock())
+    dh = JsonGzDataHandler(testdatadir)
+    assert not dh.ohlcv_purge('UNITTEST/NONEXIST', '5m', '')
+    assert not dh.ohlcv_purge('UNITTEST/NONEXIST', '5m', candle_type='mark')
+    assert unlinkmock.call_count == 0
+
+    mocker.patch.object(Path, "exists", MagicMock(return_value=True))
+    assert dh.ohlcv_purge('UNITTEST/NONEXIST', '5m', '')
+    assert dh.ohlcv_purge('UNITTEST/NONEXIST', '5m', candle_type='mark')
+    assert unlinkmock.call_count == 2
+
+
+def test_jsondatahandler_ohlcv_load(testdatadir, caplog):
+    dh = JsonDataHandler(testdatadir)
+    df = dh.ohlcv_load('XRP/ETH', '5m', 'spot')
+    assert len(df) == 711
+
+    df_mark = dh.ohlcv_load('UNITTEST/USDT', '1h', candle_type="mark")
+    assert len(df_mark) == 99
+
+    df_no_mark = dh.ohlcv_load('UNITTEST/USDT', '1h', 'spot')
+    assert len(df_no_mark) == 0
+
+    # Failure case (empty array)
+    df1 = dh.ohlcv_load('NOPAIR/XXX', '4m', 'spot')
+    assert len(df1) == 0
+    assert log_has("Could not load data for NOPAIR/XXX.", caplog)
+    assert df.columns.equals(df1.columns)
+
+
+@pytest.mark.parametrize('datahandler', ['feather', 'parquet'])
+def test_datahandler_trades_not_supported(datahandler, testdatadir, ):
+    dh = get_datahandler(testdatadir, datahandler)
+    with pytest.raises(NotImplementedError):
+        dh.trades_load('UNITTEST/ETH')
+    with pytest.raises(NotImplementedError):
+        dh.trades_store('UNITTEST/ETH', MagicMock())
+
+
+def test_jsondatahandler_trades_load(testdatadir, caplog):
+    dh = JsonGzDataHandler(testdatadir)
+    logmsg = "Old trades format detected - converting"
+    dh.trades_load('XRP/ETH')
+    assert not log_has(logmsg, caplog)
+
+    # Test conversation is happening
+    dh.trades_load('XRP/OLD')
+    assert log_has(logmsg, caplog)
+
+
+def test_jsondatahandler_trades_purge(mocker, testdatadir):
+    mocker.patch.object(Path, "exists", MagicMock(return_value=False))
+    unlinkmock = mocker.patch.object(Path, "unlink", MagicMock())
+    dh = JsonGzDataHandler(testdatadir)
+    assert not dh.trades_purge('UNITTEST/NONEXIST')
+    assert unlinkmock.call_count == 0
+
+    mocker.patch.object(Path, "exists", MagicMock(return_value=True))
+    assert dh.trades_purge('UNITTEST/NONEXIST')
+    assert unlinkmock.call_count == 1
+
+
+@pytest.mark.parametrize('datahandler', AVAILABLE_DATAHANDLERS)
+def test_datahandler_ohlcv_append(datahandler, testdatadir, ):
+    dh = get_datahandler(testdatadir, datahandler)
+    with pytest.raises(NotImplementedError):
+        dh.ohlcv_append('UNITTEST/ETH', '5m', DataFrame(), CandleType.SPOT)
+    with pytest.raises(NotImplementedError):
+        dh.ohlcv_append('UNITTEST/ETH', '5m', DataFrame(), CandleType.MARK)
+
+
+@pytest.mark.parametrize('datahandler', AVAILABLE_DATAHANDLERS)
+def test_datahandler_trades_append(datahandler, testdatadir):
+    dh = get_datahandler(testdatadir, datahandler)
+    with pytest.raises(NotImplementedError):
+        dh.trades_append('UNITTEST/ETH', [])
+
+
+def test_hdf5datahandler_trades_get_pairs(testdatadir):
+    pairs = HDF5DataHandler.trades_get_pairs(testdatadir)
+    # Convert to set to avoid failures due to sorting
+    assert set(pairs) == {'XRP/ETH'}
+
+
+def test_hdf5datahandler_trades_load(testdatadir):
+    dh = get_datahandler(testdatadir, 'hdf5')
+    trades = dh.trades_load('XRP/ETH')
+    assert isinstance(trades, list)
+
+    trades1 = dh.trades_load('UNITTEST/NONEXIST')
+    assert trades1 == []
+    # data goes from 2019-10-11 - 2019-10-13
+    timerange = TimeRange.parse_timerange('20191011-20191012')
+
+    trades2 = dh._trades_load('XRP/ETH', timerange)
+    assert len(trades) > len(trades2)
+    # Check that ID is None (If it's nan, it's wrong)
+    assert trades2[0][2] is None
+
+    # unfiltered load has trades before starttime
+    assert len([t for t in trades if t[0] < timerange.startts * 1000]) >= 0
+    # filtered list does not have trades before starttime
+    assert len([t for t in trades2 if t[0] < timerange.startts * 1000]) == 0
+    # unfiltered load has trades after endtime
+    assert len([t for t in trades if t[0] > timerange.stopts * 1000]) > 0
+    # filtered list does not have trades after endtime
+    assert len([t for t in trades2 if t[0] > timerange.stopts * 1000]) == 0
+
+
+def test_hdf5datahandler_trades_store(testdatadir, tmpdir):
+    tmpdir1 = Path(tmpdir)
+    dh = get_datahandler(testdatadir, 'hdf5')
+    trades = dh.trades_load('XRP/ETH')
+
+    dh1 = get_datahandler(tmpdir1, 'hdf5')
+    dh1.trades_store('XRP/NEW', trades)
+    file = tmpdir1 / 'XRP_NEW-trades.h5'
+    assert file.is_file()
+    # Load trades back
+    trades_new = dh1.trades_load('XRP/NEW')
+
+    assert len(trades_new) == len(trades)
+    assert trades[0][0] == trades_new[0][0]
+    assert trades[0][1] == trades_new[0][1]
+    # assert trades[0][2] == trades_new[0][2]  # This is nan - so comparison does not make sense
+    assert trades[0][3] == trades_new[0][3]
+    assert trades[0][4] == trades_new[0][4]
+    assert trades[0][5] == trades_new[0][5]
+    assert trades[0][6] == trades_new[0][6]
+    assert trades[-1][0] == trades_new[-1][0]
+    assert trades[-1][1] == trades_new[-1][1]
+    # assert trades[-1][2] == trades_new[-1][2]  # This is nan - so comparison does not make sense
+    assert trades[-1][3] == trades_new[-1][3]
+    assert trades[-1][4] == trades_new[-1][4]
+    assert trades[-1][5] == trades_new[-1][5]
+    assert trades[-1][6] == trades_new[-1][6]
+
+
+def test_hdf5datahandler_trades_purge(mocker, testdatadir):
+    mocker.patch.object(Path, "exists", MagicMock(return_value=False))
+    unlinkmock = mocker.patch.object(Path, "unlink", MagicMock())
+    dh = get_datahandler(testdatadir, 'hdf5')
+    assert not dh.trades_purge('UNITTEST/NONEXIST')
+    assert unlinkmock.call_count == 0
+
+    mocker.patch.object(Path, "exists", MagicMock(return_value=True))
+    assert dh.trades_purge('UNITTEST/NONEXIST')
+    assert unlinkmock.call_count == 1
+
+
+@pytest.mark.parametrize('pair,timeframe,candle_type,candle_append,startdt,enddt', [
+    # Data goes from 2018-01-10 - 2018-01-30
+    ('UNITTEST/BTC', '5m', 'spot',  '', '2018-01-15', '2018-01-19'),
+    # Mark data goes from to 2021-11-15 2021-11-19
+    ('UNITTEST/USDT:USDT', '1h', 'mark', '-mark', '2021-11-16', '2021-11-18'),
+])
+def test_hdf5datahandler_ohlcv_load_and_resave(
+    testdatadir,
+    tmpdir,
+    pair,
+    timeframe,
+    candle_type,
+    candle_append,
+    startdt, enddt
+):
+    tmpdir1 = Path(tmpdir)
+    tmpdir2 = tmpdir1
+    if candle_type not in ('', 'spot'):
+        tmpdir2 = tmpdir1 / 'futures'
+        tmpdir2.mkdir()
+    dh = get_datahandler(testdatadir, 'hdf5')
+    ohlcv = dh._ohlcv_load(pair, timeframe, None, candle_type=candle_type)
+    assert isinstance(ohlcv, DataFrame)
+    assert len(ohlcv) > 0
+
+    file = tmpdir2 / f"UNITTEST_NEW-{timeframe}{candle_append}.h5"
+    assert not file.is_file()
+
+    dh1 = get_datahandler(tmpdir1, 'hdf5')
+    dh1.ohlcv_store('UNITTEST/NEW', timeframe, ohlcv, candle_type=candle_type)
+    assert file.is_file()
+
+    assert not ohlcv[ohlcv['date'] < startdt].empty
+
+    timerange = TimeRange.parse_timerange(f"{startdt.replace('-', '')}-{enddt.replace('-', '')}")
+
+    # Call private function to ensure timerange is filtered in hdf5
+    ohlcv = dh._ohlcv_load(pair, timeframe, timerange, candle_type=candle_type)
+    ohlcv1 = dh1._ohlcv_load('UNITTEST/NEW', timeframe, timerange, candle_type=candle_type)
+    assert len(ohlcv) == len(ohlcv1)
+    assert ohlcv.equals(ohlcv1)
+    assert ohlcv[ohlcv['date'] < startdt].empty
+    assert ohlcv[ohlcv['date'] > enddt].empty
+
+    # Try loading inexisting file
+    ohlcv = dh.ohlcv_load('UNITTEST/NONEXIST', timeframe, candle_type=candle_type)
+    assert ohlcv.empty
+
+
+@pytest.mark.parametrize('pair,timeframe,candle_type,candle_append,startdt,enddt', [
+    # Data goes from 2018-01-10 - 2018-01-30
+    ('UNITTEST/BTC', '5m', 'spot',  '', '2018-01-15', '2018-01-19'),
+    # Mark data goes from to 2021-11-15 2021-11-19
+    ('UNITTEST/USDT', '1h', 'mark', '-mark', '2021-11-16', '2021-11-18'),
+])
+@pytest.mark.parametrize('datahandler', ['hdf5', 'feather', 'parquet'])
+def test_generic_datahandler_ohlcv_load_and_resave(
+    datahandler,
+    testdatadir,
+    tmpdir,
+    pair,
+    timeframe,
+    candle_type,
+    candle_append,
+    startdt, enddt
+):
+    tmpdir1 = Path(tmpdir)
+    tmpdir2 = tmpdir1
+    if candle_type not in ('', 'spot'):
+        tmpdir2 = tmpdir1 / 'futures'
+        tmpdir2.mkdir()
+    # Load data from one common file
+    dhbase = get_datahandler(testdatadir, 'json')
+    ohlcv = dhbase._ohlcv_load(pair, timeframe, None, candle_type=candle_type)
+    assert isinstance(ohlcv, DataFrame)
+    assert len(ohlcv) > 0
+
+    # Get data to test
+    dh = get_datahandler(testdatadir, datahandler)
+
+    file = tmpdir2 / f"UNITTEST_NEW-{timeframe}{candle_append}.{dh._get_file_extension()}"
+    assert not file.is_file()
+
+    dh1 = get_datahandler(tmpdir1, datahandler)
+    dh1.ohlcv_store('UNITTEST/NEW', timeframe, ohlcv, candle_type=candle_type)
+    assert file.is_file()
+
+    assert not ohlcv[ohlcv['date'] < startdt].empty
+
+    timerange = TimeRange.parse_timerange(f"{startdt.replace('-', '')}-{enddt.replace('-', '')}")
+
+    ohlcv = dhbase.ohlcv_load(pair, timeframe, timerange=timerange, candle_type=candle_type)
+    if datahandler == 'hdf5':
+        ohlcv1 = dh1._ohlcv_load('UNITTEST/NEW', timeframe, timerange, candle_type=candle_type)
+        if candle_type == 'mark':
+            ohlcv1['volume'] = 0.0
+    else:
+        ohlcv1 = dh1.ohlcv_load('UNITTEST/NEW', timeframe,
+                                timerange=timerange, candle_type=candle_type)
+
+    assert len(ohlcv) == len(ohlcv1)
+    assert ohlcv.equals(ohlcv1)
+    assert ohlcv[ohlcv['date'] < startdt].empty
+    assert ohlcv[ohlcv['date'] > enddt].empty
+
+    # Try loading inexisting file
+    ohlcv = dh.ohlcv_load('UNITTEST/NONEXIST', timeframe, candle_type=candle_type)
+    assert ohlcv.empty
+
+
+def test_hdf5datahandler_ohlcv_purge(mocker, testdatadir):
+    mocker.patch.object(Path, "exists", MagicMock(return_value=False))
+    unlinkmock = mocker.patch.object(Path, "unlink", MagicMock())
+    dh = get_datahandler(testdatadir, 'hdf5')
+    assert not dh.ohlcv_purge('UNITTEST/NONEXIST', '5m', '')
+    assert not dh.ohlcv_purge('UNITTEST/NONEXIST', '5m', candle_type='mark')
+    assert unlinkmock.call_count == 0
+
+    mocker.patch.object(Path, "exists", MagicMock(return_value=True))
+    assert dh.ohlcv_purge('UNITTEST/NONEXIST', '5m', '')
+    assert dh.ohlcv_purge('UNITTEST/NONEXIST', '5m', candle_type='mark')
+    assert unlinkmock.call_count == 2
+
+
+def test_gethandlerclass():
+    cl = get_datahandlerclass('json')
+    assert cl == JsonDataHandler
+    assert issubclass(cl, IDataHandler)
+
+    cl = get_datahandlerclass('jsongz')
+    assert cl == JsonGzDataHandler
+    assert issubclass(cl, IDataHandler)
+    assert issubclass(cl, JsonDataHandler)
+
+    cl = get_datahandlerclass('hdf5')
+    assert cl == HDF5DataHandler
+    assert issubclass(cl, IDataHandler)
+
+    cl = get_datahandlerclass('feather')
+    assert cl == FeatherDataHandler
+    assert issubclass(cl, IDataHandler)
+
+    cl = get_datahandlerclass('parquet')
+    assert cl == ParquetDataHandler
+    assert issubclass(cl, IDataHandler)
+
+    with pytest.raises(ValueError, match=r"No datahandler for .*"):
+        get_datahandlerclass('DeadBeef')
+
+
+def test_get_datahandler(testdatadir):
+    dh = get_datahandler(testdatadir, 'json')
+    assert type(dh) == JsonDataHandler
+    dh = get_datahandler(testdatadir, 'jsongz')
+    assert type(dh) == JsonGzDataHandler
+    dh1 = get_datahandler(testdatadir, 'jsongz', dh)
+    assert id(dh1) == id(dh)
+
+    dh = get_datahandler(testdatadir, 'hdf5')
+    assert type(dh) == HDF5DataHandler

tests/data/test_dataprovider.py

@@ -144,6 +144,77 @@ def test_available_pairs(mocker, default_conf, ohlcv_history):
     assert dp.available_pairs == [("XRP/BTC", timeframe), ("UNITTEST/BTC", timeframe), ]
 
 
+def test_producer_pairs(mocker, default_conf, ohlcv_history):
+    dataprovider = DataProvider(default_conf, None)
+
+    producer = "default"
+    whitelist = ["XRP/BTC", "ETH/BTC"]
+    assert len(dataprovider.get_producer_pairs(producer)) == 0
+
+    dataprovider._set_producer_pairs(whitelist, producer)
+    assert len(dataprovider.get_producer_pairs(producer)) == 2
+
+    new_whitelist = ["BTC/USDT"]
+    dataprovider._set_producer_pairs(new_whitelist, producer)
+    assert dataprovider.get_producer_pairs(producer) == new_whitelist
+
+    assert dataprovider.get_producer_pairs("bad") == []
+
+
+def test_get_producer_df(mocker, default_conf, ohlcv_history):
+    dataprovider = DataProvider(default_conf, None)
+
+    pair = 'BTC/USDT'
+    timeframe = default_conf['timeframe']
+    candle_type = CandleType.SPOT
+
+    empty_la = datetime.fromtimestamp(0, tz=timezone.utc)
+    now = datetime.now(timezone.utc)
+
+    # no data has been added, any request should return an empty dataframe
+    dataframe, la = dataprovider.get_producer_df(pair, timeframe, candle_type)
+    assert dataframe.empty
+    assert la == empty_la
+
+    # the data is added, should return that added dataframe
+    dataprovider._add_external_df(pair, ohlcv_history, now, timeframe, candle_type)
+    dataframe, la = dataprovider.get_producer_df(pair, timeframe, candle_type)
+    assert len(dataframe) > 0
+    assert la > empty_la
+
+    # no data on this producer, should return empty dataframe
+    dataframe, la = dataprovider.get_producer_df(pair, producer_name='bad')
+    assert dataframe.empty
+    assert la == empty_la
+
+    # non existent timeframe, empty dataframe
+    datframe, la = dataprovider.get_producer_df(pair, timeframe='1h')
+    assert dataframe.empty
+    assert la == empty_la
+
+
+def test_emit_df(mocker, default_conf, ohlcv_history):
+    mocker.patch('freqtrade.rpc.rpc_manager.RPCManager.__init__', MagicMock())
+    rpc_mock = mocker.patch('freqtrade.rpc.rpc_manager.RPCManager', MagicMock())
+    send_mock = mocker.patch('freqtrade.rpc.rpc_manager.RPCManager.send_msg', MagicMock())
+
+    dataprovider = DataProvider(default_conf, exchange=None, rpc=rpc_mock)
+    dataprovider_no_rpc = DataProvider(default_conf, exchange=None)
+
+    pair = "BTC/USDT"
+
+    # No emit yet
+    assert send_mock.call_count == 0
+
+    # Rpc is added, we call emit, should call send_msg
+    dataprovider._emit_df(pair, ohlcv_history)
+    assert send_mock.call_count == 1
+
+    # No rpc added, emit called, should not call send_msg
+    dataprovider_no_rpc._emit_df(pair, ohlcv_history)
+    assert send_mock.call_count == 1
+
+
 def test_refresh(mocker, default_conf, ohlcv_history):
     refresh_mock = MagicMock()
     mocker.patch("freqtrade.exchange.Exchange.refresh_latest_ohlcv", refresh_mock)

tests/data/test_history.py

@@ -1,7 +1,6 @@
 # pragma pylint: disable=missing-docstring, protected-access, C0103
 
 import json
-import re
 import uuid
 from pathlib import Path
 from shutil import copyfile
@@ -13,18 +12,17 @@ from pandas import DataFrame
 from pandas.testing import assert_frame_equal
 
 from freqtrade.configuration import TimeRange
-from freqtrade.constants import AVAILABLE_DATAHANDLERS, DATETIME_PRINT_FORMAT
+from freqtrade.constants import DATETIME_PRINT_FORMAT
 from freqtrade.data.converter import ohlcv_to_dataframe
-from freqtrade.data.history.hdf5datahandler import HDF5DataHandler
 from freqtrade.data.history.history_utils import (_download_pair_history, _download_trades_history,
                                                   _load_cached_data_for_updating,
                                                   convert_trades_to_ohlcv, get_timerange, load_data,
                                                   load_pair_history, refresh_backtest_ohlcv_data,
                                                   refresh_backtest_trades_data, refresh_data,
                                                   validate_backtest_data)
-from freqtrade.data.history.idatahandler import IDataHandler, get_datahandler, get_datahandlerclass
+from freqtrade.data.history.idatahandler import get_datahandler
 from freqtrade.data.history.jsondatahandler import JsonDataHandler, JsonGzDataHandler
-from freqtrade.enums import CandleType, TradingMode
+from freqtrade.enums import CandleType
 from freqtrade.exchange import timeframe_to_minutes
 from freqtrade.misc import file_dump_json
 from freqtrade.resolvers import StrategyResolver
@@ -32,25 +30,6 @@ from tests.conftest import (CURRENT_TEST_STRATEGY, get_patched_exchange, log_has
                             patch_exchange)
 
 
-# Change this if modifying UNITTEST/BTC testdatafile
-_BTC_UNITTEST_LENGTH = 13681
-
-
-def _backup_file(file: Path, copy_file: bool = False) -> None:
-    """
-    Backup existing file to avoid deleting the user file
-    :param file: complete path to the file
-    :param copy_file: keep file in place too.
-    :return: None
-    """
-    file_swp = str(file) + '.swp'
-    if file.is_file():
-        file.rename(file_swp)
-
-        if copy_file:
-            copyfile(file_swp, file)
-
-
 def _clean_test_file(file: Path) -> None:
     """
     Backup existing file to avoid deleting the user file
@@ -67,7 +46,7 @@ def _clean_test_file(file: Path) -> None:
         file_swp.rename(file)
 
 
-def test_load_data_30min_timeframe(mocker, caplog, default_conf, testdatadir) -> None:
+def test_load_data_30min_timeframe(caplog, testdatadir) -> None:
     ld = load_pair_history(pair='UNITTEST/BTC', timeframe='30m', datadir=testdatadir)
     assert isinstance(ld, DataFrame)
     assert not log_has(
@@ -76,7 +55,7 @@ def test_load_data_30min_timeframe(mocker, caplog, default_conf, testdatadir) ->
     )
 
 
-def test_load_data_7min_timeframe(mocker, caplog, default_conf, testdatadir) -> None:
+def test_load_data_7min_timeframe(caplog, testdatadir) -> None:
     ld = load_pair_history(pair='UNITTEST/BTC', timeframe='7m', datadir=testdatadir)
     assert isinstance(ld, DataFrame)
     assert ld.empty
@@ -108,7 +87,7 @@ def test_load_data_mark(ohlcv_history, mocker, caplog, testdatadir) -> None:
     )
 
 
-def test_load_data_startup_candles(mocker, caplog, default_conf, testdatadir) -> None:
+def test_load_data_startup_candles(mocker, testdatadir) -> None:
     ltfmock = mocker.patch(
         'freqtrade.data.history.jsondatahandler.JsonDataHandler._ohlcv_load',
         MagicMock(return_value=DataFrame()))
@@ -405,7 +384,7 @@ def test_load_partial_missing(testdatadir, caplog) -> None:
                    caplog)
 
 
-def test_init(default_conf, mocker) -> None:
+def test_init(default_conf) -> None:
     assert {} == load_data(
         datadir=Path(''),
         pairs=[],
@@ -685,340 +664,3 @@ def test_convert_trades_to_ohlcv(testdatadir, tmpdir, caplog):
     convert_trades_to_ohlcv(['NoDatapair'], timeframes=['1m', '5m'],
                             datadir=tmpdir1, timerange=tr, erase=True)
     assert log_has('Could not convert NoDatapair to OHLCV.', caplog)
-
-
-def test_datahandler_ohlcv_get_pairs(testdatadir):
-    pairs = JsonDataHandler.ohlcv_get_pairs(testdatadir, '5m', candle_type=CandleType.SPOT)
-    # Convert to set to avoid failures due to sorting
-    assert set(pairs) == {'UNITTEST/BTC', 'XLM/BTC', 'ETH/BTC', 'TRX/BTC', 'LTC/BTC',
-                          'XMR/BTC', 'ZEC/BTC', 'ADA/BTC', 'ETC/BTC', 'NXT/BTC',
-                          'DASH/BTC', 'XRP/ETH'}
-
-    pairs = JsonGzDataHandler.ohlcv_get_pairs(testdatadir, '8m', candle_type=CandleType.SPOT)
-    assert set(pairs) == {'UNITTEST/BTC'}
-
-    pairs = HDF5DataHandler.ohlcv_get_pairs(testdatadir, '5m', candle_type=CandleType.SPOT)
-    assert set(pairs) == {'UNITTEST/BTC'}
-
-    pairs = JsonDataHandler.ohlcv_get_pairs(testdatadir, '1h', candle_type=CandleType.MARK)
-    assert set(pairs) == {'UNITTEST/USDT', 'XRP/USDT'}
-
-    pairs = JsonGzDataHandler.ohlcv_get_pairs(testdatadir, '1h', candle_type=CandleType.FUTURES)
-    assert set(pairs) == {'XRP/USDT'}
-
-    pairs = HDF5DataHandler.ohlcv_get_pairs(testdatadir, '1h', candle_type=CandleType.MARK)
-    assert set(pairs) == {'UNITTEST/USDT:USDT'}
-
-
-@pytest.mark.parametrize('filename,pair,timeframe,candletype', [
-    ('XMR_BTC-5m.json', 'XMR_BTC', '5m', ''),
-    ('XMR_USDT-1h.h5', 'XMR_USDT', '1h', ''),
-    ('BTC-PERP-1h.h5', 'BTC-PERP', '1h', ''),
-    ('BTC_USDT-2h.jsongz', 'BTC_USDT', '2h', ''),
-    ('BTC_USDT-2h-mark.jsongz', 'BTC_USDT', '2h', 'mark'),
-    ('XMR_USDT-1h-mark.h5', 'XMR_USDT', '1h', 'mark'),
-    ('XMR_USDT-1h-random.h5', 'XMR_USDT', '1h', 'random'),
-    ('BTC-PERP-1h-index.h5', 'BTC-PERP', '1h', 'index'),
-    ('XMR_USDT_USDT-1h-mark.h5', 'XMR_USDT_USDT', '1h', 'mark'),
-])
-def test_datahandler_ohlcv_regex(filename, pair, timeframe, candletype):
-    regex = JsonDataHandler._OHLCV_REGEX
-
-    match = re.search(regex, filename)
-    assert len(match.groups()) > 1
-    assert match[1] == pair
-    assert match[2] == timeframe
-    assert match[3] == candletype
-
-
-@pytest.mark.parametrize('input,expected', [
-    ('XMR_USDT', 'XMR/USDT'),
-    ('BTC_USDT', 'BTC/USDT'),
-    ('USDT_BUSD', 'USDT/BUSD'),
-    ('BTC_USDT_USDT', 'BTC/USDT:USDT'),  # Futures
-    ('XRP_USDT_USDT', 'XRP/USDT:USDT'),  # futures
-    ('BTC-PERP', 'BTC-PERP'),
-    ('BTC-PERP_USDT', 'BTC-PERP:USDT'),  # potential FTX case
-    ('UNITTEST_USDT', 'UNITTEST/USDT'),
-])
-def test_rebuild_pair_from_filename(input, expected):
-
-    assert IDataHandler.rebuild_pair_from_filename(input) == expected
-
-
-def test_datahandler_ohlcv_get_available_data(testdatadir):
-    paircombs = JsonDataHandler.ohlcv_get_available_data(testdatadir, TradingMode.SPOT)
-    # Convert to set to avoid failures due to sorting
-    assert set(paircombs) == {
-        ('UNITTEST/BTC', '5m', CandleType.SPOT),
-        ('ETH/BTC', '5m', CandleType.SPOT),
-        ('XLM/BTC', '5m', CandleType.SPOT),
-        ('TRX/BTC', '5m', CandleType.SPOT),
-        ('LTC/BTC', '5m', CandleType.SPOT),
-        ('XMR/BTC', '5m', CandleType.SPOT),
-        ('ZEC/BTC', '5m', CandleType.SPOT),
-        ('UNITTEST/BTC', '1m', CandleType.SPOT),
-        ('ADA/BTC', '5m', CandleType.SPOT),
-        ('ETC/BTC', '5m', CandleType.SPOT),
-        ('NXT/BTC', '5m', CandleType.SPOT),
-        ('DASH/BTC', '5m', CandleType.SPOT),
-        ('XRP/ETH', '1m', CandleType.SPOT),
-        ('XRP/ETH', '5m', CandleType.SPOT),
-        ('UNITTEST/BTC', '30m', CandleType.SPOT),
-        ('UNITTEST/BTC', '8m', CandleType.SPOT),
-        ('NOPAIR/XXX', '4m', CandleType.SPOT),
-    }
-
-    paircombs = JsonDataHandler.ohlcv_get_available_data(testdatadir, TradingMode.FUTURES)
-    # Convert to set to avoid failures due to sorting
-    assert set(paircombs) == {
-        ('UNITTEST/USDT', '1h', 'mark'),
-        ('XRP/USDT', '1h', 'futures'),
-        ('XRP/USDT', '1h', 'mark'),
-        ('XRP/USDT', '8h', 'mark'),
-        ('XRP/USDT', '8h', 'funding_rate'),
-    }
-
-    paircombs = JsonGzDataHandler.ohlcv_get_available_data(testdatadir, TradingMode.SPOT)
-    assert set(paircombs) == {('UNITTEST/BTC', '8m', CandleType.SPOT)}
-    paircombs = HDF5DataHandler.ohlcv_get_available_data(testdatadir, TradingMode.SPOT)
-    assert set(paircombs) == {('UNITTEST/BTC', '5m', CandleType.SPOT)}
-
-
-def test_jsondatahandler_trades_get_pairs(testdatadir):
-    pairs = JsonGzDataHandler.trades_get_pairs(testdatadir)
-    # Convert to set to avoid failures due to sorting
-    assert set(pairs) == {'XRP/ETH', 'XRP/OLD'}
-
-
-def test_jsondatahandler_ohlcv_purge(mocker, testdatadir):
-    mocker.patch.object(Path, "exists", MagicMock(return_value=False))
-    unlinkmock = mocker.patch.object(Path, "unlink", MagicMock())
-    dh = JsonGzDataHandler(testdatadir)
-    assert not dh.ohlcv_purge('UNITTEST/NONEXIST', '5m', '')
-    assert not dh.ohlcv_purge('UNITTEST/NONEXIST', '5m', candle_type='mark')
-    assert unlinkmock.call_count == 0
-
-    mocker.patch.object(Path, "exists", MagicMock(return_value=True))
-    assert dh.ohlcv_purge('UNITTEST/NONEXIST', '5m', '')
-    assert dh.ohlcv_purge('UNITTEST/NONEXIST', '5m', candle_type='mark')
-    assert unlinkmock.call_count == 2
-
-
-def test_jsondatahandler_ohlcv_load(testdatadir, caplog):
-    dh = JsonDataHandler(testdatadir)
-    df = dh.ohlcv_load('XRP/ETH', '5m', 'spot')
-    assert len(df) == 711
-
-    df_mark = dh.ohlcv_load('UNITTEST/USDT', '1h', candle_type="mark")
-    assert len(df_mark) == 99
-
-    df_no_mark = dh.ohlcv_load('UNITTEST/USDT', '1h', 'spot')
-    assert len(df_no_mark) == 0
-
-    # Failure case (empty array)
-    df1 = dh.ohlcv_load('NOPAIR/XXX', '4m', 'spot')
-    assert len(df1) == 0
-    assert log_has("Could not load data for NOPAIR/XXX.", caplog)
-    assert df.columns.equals(df1.columns)
-
-
-def test_jsondatahandler_trades_load(testdatadir, caplog):
-    dh = JsonGzDataHandler(testdatadir)
-    logmsg = "Old trades format detected - converting"
-    dh.trades_load('XRP/ETH')
-    assert not log_has(logmsg, caplog)
-
-    # Test conversation is happening
-    dh.trades_load('XRP/OLD')
-    assert log_has(logmsg, caplog)
-
-
-def test_jsondatahandler_trades_purge(mocker, testdatadir):
-    mocker.patch.object(Path, "exists", MagicMock(return_value=False))
-    unlinkmock = mocker.patch.object(Path, "unlink", MagicMock())
-    dh = JsonGzDataHandler(testdatadir)
-    assert not dh.trades_purge('UNITTEST/NONEXIST')
-    assert unlinkmock.call_count == 0
-
-    mocker.patch.object(Path, "exists", MagicMock(return_value=True))
-    assert dh.trades_purge('UNITTEST/NONEXIST')
-    assert unlinkmock.call_count == 1
-
-
-@pytest.mark.parametrize('datahandler', AVAILABLE_DATAHANDLERS)
-def test_datahandler_ohlcv_append(datahandler, testdatadir, ):
-    dh = get_datahandler(testdatadir, datahandler)
-    with pytest.raises(NotImplementedError):
-        dh.ohlcv_append('UNITTEST/ETH', '5m', DataFrame(), CandleType.SPOT)
-    with pytest.raises(NotImplementedError):
-        dh.ohlcv_append('UNITTEST/ETH', '5m', DataFrame(), CandleType.MARK)
-
-
-@pytest.mark.parametrize('datahandler', AVAILABLE_DATAHANDLERS)
-def test_datahandler_trades_append(datahandler, testdatadir):
-    dh = get_datahandler(testdatadir, datahandler)
-    with pytest.raises(NotImplementedError):
-        dh.trades_append('UNITTEST/ETH', [])
-
-
-def test_hdf5datahandler_trades_get_pairs(testdatadir):
-    pairs = HDF5DataHandler.trades_get_pairs(testdatadir)
-    # Convert to set to avoid failures due to sorting
-    assert set(pairs) == {'XRP/ETH'}
-
-
-def test_hdf5datahandler_trades_load(testdatadir):
-    dh = HDF5DataHandler(testdatadir)
-    trades = dh.trades_load('XRP/ETH')
-    assert isinstance(trades, list)
-
-    trades1 = dh.trades_load('UNITTEST/NONEXIST')
-    assert trades1 == []
-    # data goes from 2019-10-11 - 2019-10-13
-    timerange = TimeRange.parse_timerange('20191011-20191012')
-
-    trades2 = dh._trades_load('XRP/ETH', timerange)
-    assert len(trades) > len(trades2)
-    # Check that ID is None (If it's nan, it's wrong)
-    assert trades2[0][2] is None
-
-    # unfiltered load has trades before starttime
-    assert len([t for t in trades if t[0] < timerange.startts * 1000]) >= 0
-    # filtered list does not have trades before starttime
-    assert len([t for t in trades2 if t[0] < timerange.startts * 1000]) == 0
-    # unfiltered load has trades after endtime
-    assert len([t for t in trades if t[0] > timerange.stopts * 1000]) > 0
-    # filtered list does not have trades after endtime
-    assert len([t for t in trades2 if t[0] > timerange.stopts * 1000]) == 0
-
-
-def test_hdf5datahandler_trades_store(testdatadir, tmpdir):
-    tmpdir1 = Path(tmpdir)
-    dh = HDF5DataHandler(testdatadir)
-    trades = dh.trades_load('XRP/ETH')
-
-    dh1 = HDF5DataHandler(tmpdir1)
-    dh1.trades_store('XRP/NEW', trades)
-    file = tmpdir1 / 'XRP_NEW-trades.h5'
-    assert file.is_file()
-    # Load trades back
-    trades_new = dh1.trades_load('XRP/NEW')
-
-    assert len(trades_new) == len(trades)
-    assert trades[0][0] == trades_new[0][0]
-    assert trades[0][1] == trades_new[0][1]
-    # assert trades[0][2] == trades_new[0][2]  # This is nan - so comparison does not make sense
-    assert trades[0][3] == trades_new[0][3]
-    assert trades[0][4] == trades_new[0][4]
-    assert trades[0][5] == trades_new[0][5]
-    assert trades[0][6] == trades_new[0][6]
-    assert trades[-1][0] == trades_new[-1][0]
-    assert trades[-1][1] == trades_new[-1][1]
-    # assert trades[-1][2] == trades_new[-1][2]  # This is nan - so comparison does not make sense
-    assert trades[-1][3] == trades_new[-1][3]
-    assert trades[-1][4] == trades_new[-1][4]
-    assert trades[-1][5] == trades_new[-1][5]
-    assert trades[-1][6] == trades_new[-1][6]
-
-
-def test_hdf5datahandler_trades_purge(mocker, testdatadir):
-    mocker.patch.object(Path, "exists", MagicMock(return_value=False))
-    unlinkmock = mocker.patch.object(Path, "unlink", MagicMock())
-    dh = HDF5DataHandler(testdatadir)
-    assert not dh.trades_purge('UNITTEST/NONEXIST')
-    assert unlinkmock.call_count == 0
-
-    mocker.patch.object(Path, "exists", MagicMock(return_value=True))
-    assert dh.trades_purge('UNITTEST/NONEXIST')
-    assert unlinkmock.call_count == 1
-
-
-@pytest.mark.parametrize('pair,timeframe,candle_type,candle_append,startdt,enddt', [
-    # Data goes from 2018-01-10 - 2018-01-30
-    ('UNITTEST/BTC', '5m', 'spot',  '', '2018-01-15', '2018-01-19'),
-    # Mark data goes from to 2021-11-15 2021-11-19
-    ('UNITTEST/USDT:USDT', '1h', 'mark', '-mark', '2021-11-16', '2021-11-18'),
-])
-def test_hdf5datahandler_ohlcv_load_and_resave(
-    testdatadir,
-    tmpdir,
-    pair,
-    timeframe,
-    candle_type,
-    candle_append,
-    startdt, enddt
-):
-    tmpdir1 = Path(tmpdir)
-    tmpdir2 = tmpdir1
-    if candle_type not in ('', 'spot'):
-        tmpdir2 = tmpdir1 / 'futures'
-        tmpdir2.mkdir()
-    dh = HDF5DataHandler(testdatadir)
-    ohlcv = dh._ohlcv_load(pair, timeframe, None, candle_type=candle_type)
-    assert isinstance(ohlcv, DataFrame)
-    assert len(ohlcv) > 0
-
-    file = tmpdir2 / f"UNITTEST_NEW-{timeframe}{candle_append}.h5"
-    assert not file.is_file()
-
-    dh1 = HDF5DataHandler(tmpdir1)
-    dh1.ohlcv_store('UNITTEST/NEW', timeframe, ohlcv, candle_type=candle_type)
-    assert file.is_file()
-
-    assert not ohlcv[ohlcv['date'] < startdt].empty
-
-    timerange = TimeRange.parse_timerange(f"{startdt.replace('-', '')}-{enddt.replace('-', '')}")
-
-    # Call private function to ensure timerange is filtered in hdf5
-    ohlcv = dh._ohlcv_load(pair, timeframe, timerange, candle_type=candle_type)
-    ohlcv1 = dh1._ohlcv_load('UNITTEST/NEW', timeframe, timerange, candle_type=candle_type)
-    assert len(ohlcv) == len(ohlcv1)
-    assert ohlcv.equals(ohlcv1)
-    assert ohlcv[ohlcv['date'] < startdt].empty
-    assert ohlcv[ohlcv['date'] > enddt].empty
-
-    # Try loading inexisting file
-    ohlcv = dh.ohlcv_load('UNITTEST/NONEXIST', timeframe, candle_type=candle_type)
-    assert ohlcv.empty
-
-
-def test_hdf5datahandler_ohlcv_purge(mocker, testdatadir):
-    mocker.patch.object(Path, "exists", MagicMock(return_value=False))
-    unlinkmock = mocker.patch.object(Path, "unlink", MagicMock())
-    dh = HDF5DataHandler(testdatadir)
-    assert not dh.ohlcv_purge('UNITTEST/NONEXIST', '5m', '')
-    assert not dh.ohlcv_purge('UNITTEST/NONEXIST', '5m', candle_type='mark')
-    assert unlinkmock.call_count == 0
-
-    mocker.patch.object(Path, "exists", MagicMock(return_value=True))
-    assert dh.ohlcv_purge('UNITTEST/NONEXIST', '5m', '')
-    assert dh.ohlcv_purge('UNITTEST/NONEXIST', '5m', candle_type='mark')
-    assert unlinkmock.call_count == 2
-
-
-def test_gethandlerclass():
-    cl = get_datahandlerclass('json')
-    assert cl == JsonDataHandler
-    assert issubclass(cl, IDataHandler)
-    cl = get_datahandlerclass('jsongz')
-    assert cl == JsonGzDataHandler
-    assert issubclass(cl, IDataHandler)
-    assert issubclass(cl, JsonDataHandler)
-    cl = get_datahandlerclass('hdf5')
-    assert cl == HDF5DataHandler
-    assert issubclass(cl, IDataHandler)
-    with pytest.raises(ValueError, match=r"No datahandler for .*"):
-        get_datahandlerclass('DeadBeef')
-
-
-def test_get_datahandler(testdatadir):
-    dh = get_datahandler(testdatadir, 'json')
-    assert type(dh) == JsonDataHandler
-    dh = get_datahandler(testdatadir, 'jsongz')
-    assert type(dh) == JsonGzDataHandler
-    dh1 = get_datahandler(testdatadir, 'jsongz', dh)
-    assert id(dh1) == id(dh)
-
-    dh = get_datahandler(testdatadir, 'hdf5')
-    assert type(dh) == HDF5DataHandler

tests/exchange/test_exchange.py

@@ -20,6 +20,7 @@ from freqtrade.exchange import (Binance, Bittrex, Exchange, Kraken, amount_to_pr
                                 timeframe_to_prev_date, timeframe_to_seconds)
 from freqtrade.exchange.common import (API_FETCH_ORDER_RETRY_COUNT, API_RETRY_COUNT,
                                        calculate_backoff, remove_credentials)
+from freqtrade.exchange.exchange import amount_to_contract_precision
 from freqtrade.resolvers.exchange_resolver import ExchangeResolver
 from tests.conftest import get_mock_coro, get_patched_exchange, log_has, log_has_re, num_log_has_re
 
@@ -4470,6 +4471,7 @@ def test__amount_to_contracts(
     ('ADA/USDT:USDT', 10.4445555, 10.4, 10.444),
     ('LTC/ETH', 30, 30, 30),
     ('LTC/USD', 30, 30, 30),
+    ('ADA/USDT:USDT', 1.17, 1.1, 1.17),
     # contract size of 10
     ('ETH/USDT:USDT', 10.111, 10.1, 10),
     ('ETH/USDT:USDT', 10.188, 10.1, 10),
@@ -4497,6 +4499,20 @@ def test_amount_to_contract_precision(
     assert result_size == expected_fut
 
 
+@pytest.mark.parametrize('amount,precision,precision_mode,contract_size,expected', [
+    (1.17, 1.0, 4, 0.01, 1.17),  # Tick size
+    (1.17, 1.0, 2, 0.01, 1.17),  #
+    (1.16, 1.0, 4, 0.01, 1.16),  #
+    (1.16, 1.0, 2, 0.01, 1.16),  #
+    (1.13, 1.0, 2, 0.01, 1.13),  #
+    (10.988, 1.0, 2, 10, 10),
+    (10.988, 1.0, 4, 10, 10),
+])
+def test_amount_to_contract_precision2(amount, precision, precision_mode, contract_size, expected):
+    res = amount_to_contract_precision(amount, precision, precision_mode, contract_size)
+    assert pytest.approx(res) == expected
+
+
 @pytest.mark.parametrize('exchange_name,open_rate,is_short,trading_mode,margin_mode', [
     # Bittrex
     ('bittrex', 2.0, False, 'spot', None),

tests/freqai/test_freqai_backtesting.py

@@ -3,21 +3,21 @@ from datetime import datetime, timezone
 from pathlib import Path
 from unittest.mock import PropertyMock
 
-import pytest
+from freqtrade.commands.optimize_commands import setup_optimize_configuration
-
+from freqtrade.enums import RunMode
-from freqtrade.commands.optimize_commands import start_backtesting
-from freqtrade.exceptions import OperationalException
 from freqtrade.optimize.backtesting import Backtesting
 from tests.conftest import (CURRENT_TEST_STRATEGY, get_args, log_has_re, patch_exchange,
                             patched_configuration_load_config_file)
 
 
-def test_freqai_backtest_start_backtest_list(freqai_conf, mocker, testdatadir):
+def test_freqai_backtest_start_backtest_list(freqai_conf, mocker, testdatadir, caplog):
     patch_exchange(mocker)
 
+    now = datetime.now(timezone.utc)
     mocker.patch('freqtrade.plugins.pairlistmanager.PairListManager.whitelist',
                  PropertyMock(return_value=['HULUMULU/USDT', 'XRP/USDT']))
-    # mocker.patch('freqtrade.optimize.backtesting.Backtesting.backtest', backtestmock)
+    mocker.patch('freqtrade.optimize.backtesting.history.load_data')
+    mocker.patch('freqtrade.optimize.backtesting.history.get_timerange', return_value=(now, now))
 
     patched_configuration_load_config_file(mocker, freqai_conf)
 
@@ -30,9 +30,11 @@ def test_freqai_backtest_start_backtest_list(freqai_conf, mocker, testdatadir):
         '--strategy-list', CURRENT_TEST_STRATEGY
     ]
     args = get_args(args)
-    with pytest.raises(OperationalException,
+    bt_config = setup_optimize_configuration(args, RunMode.BACKTEST)
-                       match=r"You can't use strategy_list and freqai at the same time\."):
+    Backtesting(bt_config)
-        start_backtesting(args)
+    assert log_has_re('Using --strategy-list with FreqAI REQUIRES all strategies to have identical '
+                      'populate_any_indicators.', caplog)
+    Backtesting.cleanup()
 
 
 def test_freqai_backtest_load_data(freqai_conf, mocker, caplog):

tests/optimize/test_optimize_reports.py

@@ -1,6 +1,7 @@
 import re
 from datetime import timedelta
 from pathlib import Path
+from shutil import copyfile
 
 import joblib
 import pandas as pd
@@ -25,7 +26,22 @@ from freqtrade.optimize.optimize_reports import (_get_resample_from_period, gene
                                                  text_table_exit_reason, text_table_strategy)
 from freqtrade.resolvers.strategy_resolver import StrategyResolver
 from tests.conftest import CURRENT_TEST_STRATEGY
-from tests.data.test_history import _backup_file, _clean_test_file
+from tests.data.test_history import _clean_test_file
+
+
+def _backup_file(file: Path, copy_file: bool = False) -> None:
+    """
+    Backup existing file to avoid deleting the user file
+    :param file: complete path to the file
+    :param copy_file: keep file in place too.
+    :return: None
+    """
+    file_swp = str(file) + '.swp'
+    if file.is_file():
+        file.rename(file_swp)
+
+        if copy_file:
+            copyfile(file_swp, file)
 
 
 def test_text_table_bt_results():

tests/rpc/test_rpc_apiserver.py

@@ -3,6 +3,8 @@ Unit test file for rpc/api_server.py
 """
 
 import json
+import logging
+import time
 from datetime import datetime, timedelta, timezone
 from pathlib import Path
 from unittest.mock import ANY, MagicMock, PropertyMock
@@ -10,7 +12,7 @@ from unittest.mock import ANY, MagicMock, PropertyMock
 import pandas as pd
 import pytest
 import uvicorn
-from fastapi import FastAPI
+from fastapi import FastAPI, WebSocketDisconnect
 from fastapi.exceptions import HTTPException
 from fastapi.testclient import TestClient
 from requests.auth import _basic_auth_str
@@ -31,6 +33,7 @@ from tests.conftest import (CURRENT_TEST_STRATEGY, create_mock_trades, get_mock_
 BASE_URI = "/api/v1"
 _TEST_USER = "FreqTrader"
 _TEST_PASS = "SuperSecurePassword1!"
+_TEST_WS_TOKEN = "secret_Ws_t0ken"
 
 
 @pytest.fixture
@@ -44,17 +47,21 @@ def botclient(default_conf, mocker):
                                         "CORS_origins": ['http://example.com'],
                                         "username": _TEST_USER,
                                         "password": _TEST_PASS,
+                                        "ws_token": _TEST_WS_TOKEN
                                         }})
 
     ftbot = get_patched_freqtradebot(mocker, default_conf)
     rpc = RPC(ftbot)
     mocker.patch('freqtrade.rpc.api_server.ApiServer.start_api', MagicMock())
+    apiserver = None
     try:
         apiserver = ApiServer(default_conf)
         apiserver.add_rpc_handler(rpc)
         yield ftbot, TestClient(apiserver.app)
         # Cleanup ... ?
     finally:
+        if apiserver:
+            apiserver.cleanup()
         ApiServer.shutdown()
 
 
@@ -154,6 +161,25 @@ def test_api_auth():
         get_user_from_token(b'not_a_token', 'secret1234')
 
 
+def test_api_ws_auth(botclient):
+    ftbot, client = botclient
+    def url(token): return f"/api/v1/message/ws?token={token}"
+
+    bad_token = "bad-ws_token"
+    with pytest.raises(WebSocketDisconnect):
+        with client.websocket_connect(url(bad_token)) as websocket:
+            websocket.receive()
+
+    good_token = _TEST_WS_TOKEN
+    with client.websocket_connect(url(good_token)) as websocket:
+        pass
+
+    jwt_secret = ftbot.config['api_server'].get('jwt_secret_key', 'super-secret')
+    jwt_token = create_token({'identity': {'u': 'Freqtrade'}}, jwt_secret)
+    with client.websocket_connect(url(jwt_token)) as websocket:
+        pass
+
+
 def test_api_unauthorized(botclient):
     ftbot, client = botclient
     rc = client.get(f"{BASE_URI}/ping")
@@ -261,6 +287,7 @@ def test_api__init__(default_conf, mocker):
     with pytest.raises(OperationalException, match="RPC Handler already attached."):
         apiserver.add_rpc_handler(RPC(get_patched_freqtradebot(mocker, default_conf)))
 
+    apiserver.cleanup()
     ApiServer.shutdown()
 
 
@@ -388,6 +415,7 @@ def test_api_run(default_conf, mocker, caplog):
                  MagicMock(side_effect=Exception))
     apiserver.start_api()
     assert log_has("Api server failed to start.", caplog)
+    apiserver.cleanup()
     ApiServer.shutdown()
 
 
@@ -410,6 +438,7 @@ def test_api_cleanup(default_conf, mocker, caplog):
     apiserver.cleanup()
     assert apiserver._server.cleanup.call_count == 1
     assert log_has("Stopping API Server", caplog)
+    assert log_has("Stopping API Server background tasks", caplog)
     ApiServer.shutdown()
 
 
@@ -1663,3 +1692,93 @@ def test_health(botclient):
     ret = rc.json()
     assert ret['last_process_ts'] == 0
     assert ret['last_process'] == '1970-01-01T00:00:00+00:00'
+
+
+def test_api_ws_subscribe(botclient, mocker):
+    ftbot, client = botclient
+    ws_url = f"/api/v1/message/ws?token={_TEST_WS_TOKEN}"
+
+    sub_mock = mocker.patch('freqtrade.rpc.api_server.ws.WebSocketChannel.set_subscriptions')
+
+    with client.websocket_connect(ws_url) as ws:
+        ws.send_json({'type': 'subscribe', 'data': ['whitelist']})
+
+    # Check call count is now 1 as we sent a valid subscribe request
+    assert sub_mock.call_count == 1
+
+    with client.websocket_connect(ws_url) as ws:
+        ws.send_json({'type': 'subscribe', 'data': 'whitelist'})
+
+    # Call count hasn't changed as the subscribe request was invalid
+    assert sub_mock.call_count == 1
+
+
+def test_api_ws_requests(botclient, mocker, caplog):
+    caplog.set_level(logging.DEBUG)
+
+    ftbot, client = botclient
+    ws_url = f"/api/v1/message/ws?token={_TEST_WS_TOKEN}"
+
+    # Test whitelist request
+    with client.websocket_connect(ws_url) as ws:
+        ws.send_json({"type": "whitelist", "data": None})
+        response = ws.receive_json()
+
+    assert log_has_re(r"Request of type whitelist from.+", caplog)
+    assert response['type'] == "whitelist"
+
+    # Test analyzed_df request
+    with client.websocket_connect(ws_url) as ws:
+        ws.send_json({"type": "analyzed_df", "data": {}})
+        response = ws.receive_json()
+
+    assert log_has_re(r"Request of type analyzed_df from.+", caplog)
+    assert response['type'] == "analyzed_df"
+
+    caplog.clear()
+    # Test analyzed_df request with data
+    with client.websocket_connect(ws_url) as ws:
+        ws.send_json({"type": "analyzed_df", "data": {"limit": 100}})
+        response = ws.receive_json()
+
+    assert log_has_re(r"Request of type analyzed_df from.+", caplog)
+    assert response['type'] == "analyzed_df"
+
+
+def test_api_ws_send_msg(default_conf, mocker, caplog):
+    try:
+        caplog.set_level(logging.DEBUG)
+
+        default_conf.update({"api_server": {"enabled": True,
+                                            "listen_ip_address": "127.0.0.1",
+                                            "listen_port": 8080,
+                                            "CORS_origins": ['http://example.com'],
+                                            "username": _TEST_USER,
+                                            "password": _TEST_PASS,
+                                            "ws_token": _TEST_WS_TOKEN
+                                            }})
+        mocker.patch('freqtrade.rpc.telegram.Updater')
+        mocker.patch('freqtrade.rpc.api_server.ApiServer.start_api')
+        apiserver = ApiServer(default_conf)
+        apiserver.add_rpc_handler(RPC(get_patched_freqtradebot(mocker, default_conf)))
+        apiserver.start_message_queue()
+        # Give the queue thread time to start
+        time.sleep(0.2)
+
+        # Test message_queue coro receives the message
+        test_message = {"type": "status", "data": "test"}
+        apiserver.send_msg(test_message)
+        time.sleep(0.1)  # Not sure how else to wait for the coro to receive the data
+        assert log_has("Found message of type: status", caplog)
+
+        # Test if exception logged when error occurs in sending
+        mocker.patch('freqtrade.rpc.api_server.ws.channel.ChannelManager.broadcast',
+                     side_effect=Exception)
+
+        apiserver.send_msg(test_message)
+        time.sleep(0.1)  # Not sure how else to wait for the coro to receive the data
+        assert log_has_re(r"Exception happened in background task.*", caplog)
+
+    finally:
+        apiserver.cleanup()
+        ApiServer.shutdown()

tests/rpc/test_rpc_emc.py

@@ -0,0 +1,465 @@
+"""
+Unit test file for rpc/external_message_consumer.py
+"""
+import asyncio
+import functools
+import logging
+from datetime import datetime, timezone
+from unittest.mock import MagicMock
+
+import pytest
+import websockets
+
+from freqtrade.data.dataprovider import DataProvider
+from freqtrade.rpc.external_message_consumer import ExternalMessageConsumer
+from tests.conftest import log_has, log_has_re, log_has_when
+
+
+_TEST_WS_TOKEN = "secret_Ws_t0ken"
+_TEST_WS_HOST = "127.0.0.1"
+_TEST_WS_PORT = 9989
+
+
+@pytest.fixture
+def patched_emc(default_conf, mocker):
+    default_conf.update({
+        "external_message_consumer": {
+            "enabled": True,
+            "producers": [
+                {
+                    "name": "default",
+                    "host": "null",
+                    "port": 9891,
+                    "ws_token": _TEST_WS_TOKEN
+                }
+            ]
+        }
+    })
+    dataprovider = DataProvider(default_conf, None, None, None)
+    emc = ExternalMessageConsumer(default_conf, dataprovider)
+
+    try:
+        yield emc
+    finally:
+        emc.shutdown()
+
+
+def test_emc_start(patched_emc, caplog):
+    # Test if the message was printed
+    assert log_has_when("Starting ExternalMessageConsumer", caplog, "setup")
+    # Test if the thread and loop objects were created
+    assert patched_emc._thread and patched_emc._loop
+
+    # Test we call start again nothing happens
+    prev_thread = patched_emc._thread
+    patched_emc.start()
+    assert prev_thread == patched_emc._thread
+
+
+def test_emc_shutdown(patched_emc, caplog):
+    patched_emc.shutdown()
+
+    assert log_has("Stopping ExternalMessageConsumer", caplog)
+    # Test the loop has stopped
+    assert patched_emc._loop is None
+    # Test if the thread has stopped
+    assert patched_emc._thread is None
+
+    caplog.clear()
+    patched_emc.shutdown()
+
+    # Test func didn't run again as it was called once already
+    assert not log_has("Stopping ExternalMessageConsumer", caplog)
+
+
+def test_emc_init(patched_emc):
+    # Test the settings were set correctly
+    assert patched_emc.initial_candle_limit <= 1500
+    assert patched_emc.wait_timeout > 0
+    assert patched_emc.sleep_time > 0
+
+
+# Parametrize this?
+def test_emc_handle_producer_message(patched_emc, caplog, ohlcv_history):
+    test_producer = {"name": "test", "url": "ws://test", "ws_token": "test"}
+    producer_name = test_producer['name']
+
+    caplog.set_level(logging.DEBUG)
+
+    # Test handle whitelist message
+    whitelist_message = {"type": "whitelist", "data": ["BTC/USDT"]}
+    patched_emc.handle_producer_message(test_producer, whitelist_message)
+
+    assert log_has(f"Received message of type `whitelist` from `{producer_name}`", caplog)
+    assert log_has(
+        f"Consumed message from `{producer_name}` of type `RPCMessageType.WHITELIST`", caplog)
+
+    # Test handle analyzed_df message
+    df_message = {
+        "type": "analyzed_df",
+        "data": {
+            "key": ("BTC/USDT", "5m", "spot"),
+            "df": ohlcv_history,
+            "la": datetime.now(timezone.utc)
+        }
+    }
+    patched_emc.handle_producer_message(test_producer, df_message)
+
+    assert log_has(f"Received message of type `analyzed_df` from `{producer_name}`", caplog)
+    assert log_has(
+        f"Consumed message from `{producer_name}` of type `RPCMessageType.ANALYZED_DF`", caplog)
+
+    # Test unhandled message
+    unhandled_message = {"type": "status", "data": "RUNNING"}
+    patched_emc.handle_producer_message(test_producer, unhandled_message)
+
+    assert log_has_re(r"Received unhandled message\: .*", caplog)
+
+    # Test malformed messages
+    caplog.clear()
+    malformed_message = {"type": "whitelist", "data": {"pair": "BTC/USDT"}}
+    patched_emc.handle_producer_message(test_producer, malformed_message)
+
+    assert log_has_re(r"Invalid message .+", caplog)
+
+    malformed_message = {
+        "type": "analyzed_df",
+        "data": {
+            "key": "BTC/USDT",
+            "df": ohlcv_history,
+            "la": datetime.now(timezone.utc)
+        }
+    }
+    patched_emc.handle_producer_message(test_producer, malformed_message)
+
+    assert log_has(f"Received message of type `analyzed_df` from `{producer_name}`", caplog)
+    assert log_has_re(r"Invalid message .+", caplog)
+
+    caplog.clear()
+    malformed_message = {"some": "stuff"}
+    patched_emc.handle_producer_message(test_producer, malformed_message)
+
+    assert log_has_re(r"Invalid message .+", caplog)
+
+    caplog.clear()
+    malformed_message = {"type": "whitelist", "data": None}
+    patched_emc.handle_producer_message(test_producer, malformed_message)
+
+    assert log_has_re(r"Empty message .+", caplog)
+
+
+async def test_emc_create_connection_success(default_conf, caplog, mocker):
+    default_conf.update({
+        "external_message_consumer": {
+            "enabled": True,
+            "producers": [
+                {
+                    "name": "default",
+                    "host": _TEST_WS_HOST,
+                    "port": _TEST_WS_PORT,
+                    "ws_token": _TEST_WS_TOKEN
+                }
+            ],
+            "wait_timeout": 60,
+            "ping_timeout": 60,
+            "sleep_timeout": 60
+        }
+    })
+
+    mocker.patch('freqtrade.rpc.external_message_consumer.ExternalMessageConsumer.start',
+                 MagicMock())
+    dp = DataProvider(default_conf, None, None, None)
+    emc = ExternalMessageConsumer(default_conf, dp)
+
+    test_producer = default_conf['external_message_consumer']['producers'][0]
+    lock = asyncio.Lock()
+
+    emc._running = True
+
+    async def eat(websocket):
+        emc._running = False
+
+    try:
+        async with websockets.serve(eat, _TEST_WS_HOST, _TEST_WS_PORT):
+            await emc._create_connection(test_producer, lock)
+
+        assert log_has_re(r"Producer connection success.+", caplog)
+    finally:
+        emc.shutdown()
+
+
+async def test_emc_create_connection_invalid_port(default_conf, caplog, mocker):
+    default_conf.update({
+        "external_message_consumer": {
+            "enabled": True,
+            "producers": [
+                {
+                    "name": "default",
+                    "host": _TEST_WS_HOST,
+                    "port": -1,
+                    "ws_token": _TEST_WS_TOKEN
+                }
+            ],
+            "wait_timeout": 60,
+            "ping_timeout": 60,
+            "sleep_timeout": 60
+        }
+    })
+
+    dp = DataProvider(default_conf, None, None, None)
+    emc = ExternalMessageConsumer(default_conf, dp)
+
+    try:
+        await asyncio.sleep(0.01)
+        assert log_has_re(r".+ is an invalid WebSocket URL .+", caplog)
+    finally:
+        emc.shutdown()
+
+
+async def test_emc_create_connection_invalid_host(default_conf, caplog, mocker):
+    default_conf.update({
+        "external_message_consumer": {
+            "enabled": True,
+            "producers": [
+                {
+                    "name": "default",
+                    "host": "10000.1241..2121/",
+                    "port": _TEST_WS_PORT,
+                    "ws_token": _TEST_WS_TOKEN
+                }
+            ],
+            "wait_timeout": 60,
+            "ping_timeout": 60,
+            "sleep_timeout": 60
+        }
+    })
+
+    dp = DataProvider(default_conf, None, None, None)
+    emc = ExternalMessageConsumer(default_conf, dp)
+
+    try:
+        await asyncio.sleep(0.01)
+        assert log_has_re(r".+ is an invalid WebSocket URL .+", caplog)
+    finally:
+        emc.shutdown()
+
+
+async def test_emc_create_connection_error(default_conf, caplog, mocker):
+    default_conf.update({
+        "external_message_consumer": {
+            "enabled": True,
+            "producers": [
+                {
+                    "name": "default",
+                    "host": _TEST_WS_HOST,
+                    "port": _TEST_WS_PORT,
+                    "ws_token": _TEST_WS_TOKEN
+                }
+            ],
+            "wait_timeout": 60,
+            "ping_timeout": 60,
+            "sleep_timeout": 60
+        }
+    })
+
+    # Test unexpected error
+    mocker.patch('websockets.connect', side_effect=RuntimeError)
+
+    dp = DataProvider(default_conf, None, None, None)
+    emc = ExternalMessageConsumer(default_conf, dp)
+
+    try:
+        await asyncio.sleep(0.01)
+        assert log_has("Unexpected error has occurred:", caplog)
+    finally:
+        emc.shutdown()
+
+
+async def test_emc_receive_messages_valid(default_conf, caplog, mocker):
+    default_conf.update({
+        "external_message_consumer": {
+            "enabled": True,
+            "producers": [
+                {
+                    "name": "default",
+                    "host": _TEST_WS_HOST,
+                    "port": _TEST_WS_PORT,
+                    "ws_token": _TEST_WS_TOKEN
+                }
+            ],
+            "wait_timeout": 1,
+            "ping_timeout": 60,
+            "sleep_time": 60
+        }
+    })
+
+    mocker.patch('freqtrade.rpc.external_message_consumer.ExternalMessageConsumer.start',
+                 MagicMock())
+
+    lock = asyncio.Lock()
+    test_producer = default_conf['external_message_consumer']['producers'][0]
+
+    dp = DataProvider(default_conf, None, None, None)
+    emc = ExternalMessageConsumer(default_conf, dp)
+
+    loop = asyncio.get_event_loop()
+    def change_running(emc): emc._running = not emc._running
+
+    class TestChannel:
+        async def recv(self, *args, **kwargs):
+            return {"type": "whitelist", "data": ["BTC/USDT"]}
+
+        async def ping(self, *args, **kwargs):
+            return asyncio.Future()
+
+    try:
+        change_running(emc)
+        loop.call_soon(functools.partial(change_running, emc=emc))
+        await emc._receive_messages(TestChannel(), test_producer, lock)
+
+        assert log_has_re(r"Received message of type `whitelist`.+", caplog)
+    finally:
+        emc.shutdown()
+
+
+async def test_emc_receive_messages_invalid(default_conf, caplog, mocker):
+    default_conf.update({
+        "external_message_consumer": {
+            "enabled": True,
+            "producers": [
+                {
+                    "name": "default",
+                    "host": _TEST_WS_HOST,
+                    "port": _TEST_WS_PORT,
+                    "ws_token": _TEST_WS_TOKEN
+                }
+            ],
+            "wait_timeout": 1,
+            "ping_timeout": 60,
+            "sleep_time": 60
+        }
+    })
+
+    mocker.patch('freqtrade.rpc.external_message_consumer.ExternalMessageConsumer.start',
+                 MagicMock())
+
+    lock = asyncio.Lock()
+    test_producer = default_conf['external_message_consumer']['producers'][0]
+
+    dp = DataProvider(default_conf, None, None, None)
+    emc = ExternalMessageConsumer(default_conf, dp)
+
+    loop = asyncio.get_event_loop()
+    def change_running(emc): emc._running = not emc._running
+
+    class TestChannel:
+        async def recv(self, *args, **kwargs):
+            return {"type": ["BTC/USDT"]}
+
+        async def ping(self, *args, **kwargs):
+            return asyncio.Future()
+
+    try:
+        change_running(emc)
+        loop.call_soon(functools.partial(change_running, emc=emc))
+        await emc._receive_messages(TestChannel(), test_producer, lock)
+
+        assert log_has_re(r"Invalid message from.+", caplog)
+    finally:
+        emc.shutdown()
+
+
+async def test_emc_receive_messages_timeout(default_conf, caplog, mocker):
+    default_conf.update({
+        "external_message_consumer": {
+            "enabled": True,
+            "producers": [
+                {
+                    "name": "default",
+                    "host": _TEST_WS_HOST,
+                    "port": _TEST_WS_PORT,
+                    "ws_token": _TEST_WS_TOKEN
+                }
+            ],
+            "wait_timeout": 0.1,
+            "ping_timeout": 1,
+            "sleep_time": 1
+        }
+    })
+
+    mocker.patch('freqtrade.rpc.external_message_consumer.ExternalMessageConsumer.start',
+                 MagicMock())
+
+    lock = asyncio.Lock()
+    test_producer = default_conf['external_message_consumer']['producers'][0]
+
+    dp = DataProvider(default_conf, None, None, None)
+    emc = ExternalMessageConsumer(default_conf, dp)
+
+    loop = asyncio.get_event_loop()
+    def change_running(emc): emc._running = not emc._running
+
+    class TestChannel:
+        async def recv(self, *args, **kwargs):
+            await asyncio.sleep(0.2)
+
+        async def ping(self, *args, **kwargs):
+            return asyncio.Future()
+
+    try:
+        change_running(emc)
+        loop.call_soon(functools.partial(change_running, emc=emc))
+        await emc._receive_messages(TestChannel(), test_producer, lock)
+
+        assert log_has_re(r"Ping error.+", caplog)
+    finally:
+        emc.shutdown()
+
+
+async def test_emc_receive_messages_handle_error(default_conf, caplog, mocker):
+    default_conf.update({
+        "external_message_consumer": {
+            "enabled": True,
+            "producers": [
+                {
+                    "name": "default",
+                    "host": _TEST_WS_HOST,
+                    "port": _TEST_WS_PORT,
+                    "ws_token": _TEST_WS_TOKEN
+                }
+            ],
+            "wait_timeout": 1,
+            "ping_timeout": 1,
+            "sleep_time": 1
+        }
+    })
+
+    mocker.patch('freqtrade.rpc.external_message_consumer.ExternalMessageConsumer.start',
+                 MagicMock())
+
+    lock = asyncio.Lock()
+    test_producer = default_conf['external_message_consumer']['producers'][0]
+
+    dp = DataProvider(default_conf, None, None, None)
+    emc = ExternalMessageConsumer(default_conf, dp)
+
+    emc.handle_producer_message = MagicMock(side_effect=Exception)
+
+    loop = asyncio.get_event_loop()
+    def change_running(emc): emc._running = not emc._running
+
+    class TestChannel:
+        async def recv(self, *args, **kwargs):
+            return {"type": "whitelist", "data": ["BTC/USDT"]}
+
+        async def ping(self, *args, **kwargs):
+            return asyncio.Future()
+
+    try:
+        change_running(emc)
+        loop.call_soon(functools.partial(change_running, emc=emc))
+        await emc._receive_messages(TestChannel(), test_producer, lock)
+
+        assert log_has_re(r"Error handling producer message.+", caplog)
+    finally:
+        emc.shutdown()

tests/test_configuration.py

@@ -1089,6 +1089,58 @@ def test__validate_pricing_rules(default_conf, caplog) -> None:
         validate_config_consistency(conf)
 
 
+def test__validate_consumers(default_conf, caplog) -> None:
+    conf = deepcopy(default_conf)
+    conf.update({
+            "external_message_consumer": {
+                "enabled": True,
+                "producers": []
+                }
+            })
+    with pytest.raises(OperationalException,
+                       match="You must specify at least 1 Producer to connect to."):
+        validate_config_consistency(conf)
+
+    conf = deepcopy(default_conf)
+    conf.update({
+        "external_message_consumer": {
+            "enabled": True,
+            "producers": [
+                {
+                    "name": "default",
+                    "host": "127.0.0.1",
+                    "port": 8081,
+                    "ws_token": "secret_ws_t0ken."
+                }, {
+                    "name": "default",
+                    "host": "127.0.0.1",
+                    "port": 8080,
+                    "ws_token": "secret_ws_t0ken."
+                }
+            ]}
+        })
+    with pytest.raises(OperationalException,
+                       match="Producer names must be unique. Duplicate: default"):
+        validate_config_consistency(conf)
+
+    conf = deepcopy(default_conf)
+    conf.update({
+        "process_only_new_candles": True,
+        "external_message_consumer": {
+            "enabled": True,
+            "producers": [
+                {
+                    "name": "default",
+                    "host": "127.0.0.1",
+                    "port": 8081,
+                    "ws_token": "secret_ws_t0ken."
+                }
+            ]}
+        })
+    validate_config_consistency(conf)
+    assert log_has_re("To receive best performance with external data.*", caplog)
+
+
 def test_load_config_test_comments() -> None:
     """
     Load config with comments

tests/test_freqtradebot.py

@@ -1319,9 +1319,9 @@ def test_create_stoploss_order_invalid_order(
     assert create_order_mock.call_args[1]['amount'] == trade.amount
 
     # Rpc is sending first buy, then sell
-    assert rpc_mock.call_count == 2
+    assert rpc_mock.call_count == 3
-    assert rpc_mock.call_args_list[1][0][0]['sell_reason'] == ExitType.EMERGENCY_EXIT.value
+    assert rpc_mock.call_args_list[2][0][0]['sell_reason'] == ExitType.EMERGENCY_EXIT.value
-    assert rpc_mock.call_args_list[1][0][0]['order_type'] == 'market'
+    assert rpc_mock.call_args_list[2][0][0]['order_type'] == 'market'
 
 
 @pytest.mark.parametrize("is_short", [False, True])
@@ -2439,7 +2439,7 @@ def test_manage_open_orders_entry_usercustom(
     # Trade should be closed since the function returns true
     freqtrade.manage_open_orders()
     assert cancel_order_wr_mock.call_count == 1
-    assert rpc_mock.call_count == 1
+    assert rpc_mock.call_count == 2
     trades = Trade.query.filter(Trade.open_order_id.is_(open_trade.open_order_id)).all()
     nb_trades = len(trades)
     assert nb_trades == 0
@@ -2478,7 +2478,7 @@ def test_manage_open_orders_entry(
     # check it does cancel buy orders over the time limit
     freqtrade.manage_open_orders()
     assert cancel_order_mock.call_count == 1
-    assert rpc_mock.call_count == 1
+    assert rpc_mock.call_count == 2
     trades = Trade.query.filter(Trade.open_order_id.is_(open_trade.open_order_id)).all()
     nb_trades = len(trades)
     assert nb_trades == 0
@@ -2608,7 +2608,7 @@ def test_check_handle_cancelled_buy(
     # check it does cancel buy orders over the time limit
     freqtrade.manage_open_orders()
     assert cancel_order_mock.call_count == 0
-    assert rpc_mock.call_count == 1
+    assert rpc_mock.call_count == 2
     trades = Trade.query.filter(Trade.open_order_id.is_(open_trade.open_order_id)).all()
     assert len(trades) == 0
     assert log_has_re(
@@ -2639,7 +2639,7 @@ def test_manage_open_orders_buy_exception(
     # check it does cancel buy orders over the time limit
     freqtrade.manage_open_orders()
     assert cancel_order_mock.call_count == 0
-    assert rpc_mock.call_count == 0
+    assert rpc_mock.call_count == 1
     trades = Trade.query.filter(Trade.open_order_id.is_(open_trade.open_order_id)).all()
     nb_trades = len(trades)
     assert nb_trades == 1
@@ -2686,7 +2686,7 @@ def test_manage_open_orders_exit_usercustom(
     # Return false - No impact
     freqtrade.manage_open_orders()
     assert cancel_order_mock.call_count == 0
-    assert rpc_mock.call_count == 0
+    assert rpc_mock.call_count == 1
     assert open_trade_usdt.is_open is False
     assert freqtrade.strategy.check_exit_timeout.call_count == 1
     assert freqtrade.strategy.check_entry_timeout.call_count == 0
@@ -2696,7 +2696,7 @@ def test_manage_open_orders_exit_usercustom(
     # Return Error - No impact
     freqtrade.manage_open_orders()
     assert cancel_order_mock.call_count == 0
-    assert rpc_mock.call_count == 0
+    assert rpc_mock.call_count == 1
     assert open_trade_usdt.is_open is False
     assert freqtrade.strategy.check_exit_timeout.call_count == 1
     assert freqtrade.strategy.check_entry_timeout.call_count == 0
@@ -2706,7 +2706,7 @@ def test_manage_open_orders_exit_usercustom(
     freqtrade.strategy.check_entry_timeout = MagicMock(return_value=True)
     freqtrade.manage_open_orders()
     assert cancel_order_mock.call_count == 1
-    assert rpc_mock.call_count == 1
+    assert rpc_mock.call_count == 2
     assert open_trade_usdt.is_open is True
     assert freqtrade.strategy.check_exit_timeout.call_count == 1
     assert freqtrade.strategy.check_entry_timeout.call_count == 0
@@ -2766,7 +2766,7 @@ def test_manage_open_orders_exit(
     # check it does cancel sell orders over the time limit
     freqtrade.manage_open_orders()
     assert cancel_order_mock.call_count == 1
-    assert rpc_mock.call_count == 1
+    assert rpc_mock.call_count == 2
     assert open_trade_usdt.is_open is True
     # Custom user sell-timeout is never called
     assert freqtrade.strategy.check_exit_timeout.call_count == 0
@@ -2805,7 +2805,7 @@ def test_check_handle_cancelled_exit(
     # check it does cancel sell orders over the time limit
     freqtrade.manage_open_orders()
     assert cancel_order_mock.call_count == 0
-    assert rpc_mock.call_count == 1
+    assert rpc_mock.call_count == 2
     assert open_trade_usdt.is_open is True
     exit_name = 'Buy' if is_short else 'Sell'
     assert log_has_re(f"{exit_name} order cancelled on exchange for Trade.*", caplog)
@@ -2843,7 +2843,7 @@ def test_manage_open_orders_partial(
     # note this is for a partially-complete buy order
     freqtrade.manage_open_orders()
     assert cancel_order_mock.call_count == 1
-    assert rpc_mock.call_count == 2
+    assert rpc_mock.call_count == 3
     trades = Trade.query.filter(Trade.open_order_id.is_(open_trade.open_order_id)).all()
     assert len(trades) == 1
     assert trades[0].amount == 23.0
@@ -2890,7 +2890,7 @@ def test_manage_open_orders_partial_fee(
     assert log_has_re(r"Applying fee on amount for Trade.*", caplog)
 
     assert cancel_order_mock.call_count == 1
-    assert rpc_mock.call_count == 2
+    assert rpc_mock.call_count == 3
     trades = Trade.query.filter(Trade.open_order_id.is_(open_trade.open_order_id)).all()
     assert len(trades) == 1
     # Verify that trade has been updated
@@ -2940,7 +2940,7 @@ def test_manage_open_orders_partial_except(
     assert log_has_re(r"Could not update trade amount: .*", caplog)
 
     assert cancel_order_mock.call_count == 1
-    assert rpc_mock.call_count == 2
+    assert rpc_mock.call_count == 3
     trades = Trade.query.filter(Trade.open_order_id.is_(open_trade.open_order_id)).all()
     assert len(trades) == 1
     # Verify that trade has been updated
@@ -3155,7 +3155,7 @@ def test_handle_cancel_exit_limit(mocker, default_conf_usdt, fee) -> None:
     reason = CANCEL_REASON['TIMEOUT']
     assert freqtrade.handle_cancel_exit(trade, order, reason)
     assert cancel_order_mock.call_count == 1
-    assert send_msg_mock.call_count == 1
+    assert send_msg_mock.call_count == 2
     assert trade.close_rate is None
     assert trade.exit_reason is None
 
@@ -3592,7 +3592,7 @@ def test_execute_trade_exit_with_stoploss_on_exchange(
     trade.is_short = is_short
     assert trade
     assert cancel_order.call_count == 1
-    assert rpc_mock.call_count == 3
+    assert rpc_mock.call_count == 4
 
 
 @pytest.mark.parametrize("is_short", [False, True])
@@ -3662,11 +3662,11 @@ def test_may_execute_trade_exit_after_stoploss_on_exchange_hit(
     assert trade.stoploss_order_id is None
     assert trade.is_open is False
     assert trade.exit_reason == ExitType.STOPLOSS_ON_EXCHANGE.value
-    assert rpc_mock.call_count == 3
+    assert rpc_mock.call_count == 4
-    assert rpc_mock.call_args_list[0][0][0]['type'] == RPCMessageType.ENTRY
+    assert rpc_mock.call_args_list[1][0][0]['type'] == RPCMessageType.ENTRY
-    assert rpc_mock.call_args_list[0][0][0]['amount'] > 20
+    assert rpc_mock.call_args_list[1][0][0]['amount'] > 20
-    assert rpc_mock.call_args_list[1][0][0]['type'] == RPCMessageType.ENTRY_FILL
+    assert rpc_mock.call_args_list[2][0][0]['type'] == RPCMessageType.ENTRY_FILL
-    assert rpc_mock.call_args_list[2][0][0]['type'] == RPCMessageType.EXIT_FILL
+    assert rpc_mock.call_args_list[3][0][0]['type'] == RPCMessageType.EXIT_FILL
 
 
 @pytest.mark.parametrize(

tests/test_misc.py

@@ -7,10 +7,11 @@ from unittest.mock import MagicMock
 
 import pytest
 
-from freqtrade.misc import (decimals_per_coin, deep_merge_dicts, file_dump_json, file_load_json,
+from freqtrade.misc import (dataframe_to_json, decimals_per_coin, deep_merge_dicts, file_dump_json,
-                            format_ms_time, pair_to_filename, parse_db_uri_for_logging, plural,
+                            file_load_json, format_ms_time, json_to_dataframe, pair_to_filename,
-                            render_template, render_template_with_fallback, round_coin_value,
+                            parse_db_uri_for_logging, plural, render_template,
-                            safe_value_fallback, safe_value_fallback2, shorten_date)
+                            render_template_with_fallback, round_coin_value, safe_value_fallback,
+                            safe_value_fallback2, shorten_date)
 
 
 def test_decimals_per_coin():
@@ -219,3 +220,14 @@ def test_deep_merge_dicts():
 
     res2['first']['rows']['test'] = 'asdf'
     assert deep_merge_dicts(a, deepcopy(b), allow_null_overrides=False) == res2
+
+
+def test_dataframe_json(ohlcv_history):
+    from pandas.testing import assert_frame_equal
+    json = dataframe_to_json(ohlcv_history)
+    dataframe = json_to_dataframe(json)
+
+    assert list(ohlcv_history.columns) == list(dataframe.columns)
+    assert len(ohlcv_history) == len(dataframe)
+
+    assert_frame_equal(ohlcv_history, dataframe)