adelorenzo/stable
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Merge remote-tracking branch 'origin/develop' into feat/add-pytorch-model-support

Browse Source
This commit is contained in:
robcaulk
2023-04-08 13:22:25 +02:00
parent 48d3c8e62e f8d89c46e5
commit 69b9b35a08
123 changed files with 7298 additions and 3422 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
docs/bot-basics.md
View File

@@ -12,6 +12,9 @@ This page provides you some basic concepts on how Freqtrade works and operates.
* **Indicators**: Technical indicators (SMA, EMA, RSI, ...).
* **Limit order**: Limit orders which execute at the defined limit price or better.
* **Market order**: Guaranteed to fill, may move price depending on the order size.
* **Current Profit**: Currently pending (unrealized) profit for this trade. This is mainly used throughout the bot and UI.
* **Realized Profit**: Already realized profit. Only relevant in combination with [partial exits](strategy-callbacks.md#adjust-trade-position) - which also explains the calculation logic for this.
* **Total Profit**: Combined realized and unrealized profit. The relative number (%) is calculated against the total investment in this trade.
## Fee handling
@@ -57,10 +60,10 @@ This loop will be repeated again and again until the bot is stopped.
* Load historic data for configured pairlist.
* Calls `bot_start()` once.
* Calls `bot_loop_start()` once.
* Calculate indicators (calls `populate_indicators()` once per pair).
* Calculate entry / exit signals (calls `populate_entry_trend()` and `populate_exit_trend()` once per pair).
* Loops per candle simulating entry and exit points.
* Calls `bot_loop_start()` strategy callback.
* Check for Order timeouts, either via the `unfilledtimeout` configuration, or via `check_entry_timeout()` / `check_exit_timeout()` strategy callbacks.
* Calls `adjust_entry_price()` strategy callback for open entry orders.
* Check for trade entry signals (`enter_long` / `enter_short` columns).

8
docs/freqai-feature-engineering.md
View File

@@ -6,8 +6,8 @@ Low level feature engineering is performed in the user strategy within a set of
| Function | Description |
|---------------|-------------|
| `feature_engineering__expand_all()` | This optional function will automatically expand the defined features on the config defined `indicator_periods_candles`, `include_timeframes`, `include_shifted_candles`, and `include_corr_pairs`.
| `feature_engineering__expand_basic()` | This optional function will automatically expand the defined features on the config defined `include_timeframes`, `include_shifted_candles`, and `include_corr_pairs`. Note: this function does *not* expand across `include_periods_candles`.
| `feature_engineering_expand_all()` | This optional function will automatically expand the defined features on the config defined `indicator_periods_candles`, `include_timeframes`, `include_shifted_candles`, and `include_corr_pairs`.
| `feature_engineering_expand_basic()` | This optional function will automatically expand the defined features on the config defined `include_timeframes`, `include_shifted_candles`, and `include_corr_pairs`. Note: this function does *not* expand across `include_periods_candles`.
| `feature_engineering_standard()` | This optional function will be called once with the dataframe of the base timeframe. This is the final function to be called, which means that the dataframe entering this function will contain all the features and columns from the base asset created by the other `feature_engineering_expand` functions. This function is a good place to do custom exotic feature extractions (e.g. tsfresh). This function is also a good place for any feature that should not be auto-expanded upon (e.g., day of the week).
| `set_freqai_targets()` | Required function to set the targets for the model. All targets must be prepended with `&` to be recognized by the FreqAI internals.
@@ -182,11 +182,11 @@ In total, the number of features the user of the presented example strat has cre
$= 3 * 3 * 3 * 2 * 2 = 108$.
### Gain finer control over `feature_engineering_*` functions with `metadata`
### Gain finer control over `feature_engineering_*` functions with `metadata`
All `feature_engineering_*` and `set_freqai_targets()` functions are passed a `metadata` dictionary which contains information about the `pair`, `tf` (timeframe), and `period` that FreqAI is automating for feature building. As such, a user can use `metadata` inside `feature_engineering_*` functions as criteria for blocking/reserving features for certain timeframes, periods, pairs etc.
```py
```python
def feature_engineering_expand_all(self, dataframe, period, metadata, **kwargs):
if metadata["tf"] == "1h":
dataframe["%-roc-period"] = ta.ROC(dataframe, timeperiod=period)

3
docs/freqai-parameter-table.md
View File

@@ -46,7 +46,7 @@ Mandatory parameters are marked as **Required** and have to be set in one of the
| `outlier_protection_percentage` | Enable to prevent outlier detection methods from discarding too much data. If more than `outlier_protection_percentage` % of points are detected as outliers by the SVM or DBSCAN, FreqAI will log a warning message and ignore outlier detection, i.e., the original dataset will be kept intact. If the outlier protection is triggered, no predictions will be made based on the training dataset. <br> **Datatype:** Float. <br> Default: `30`.
| `reverse_train_test_order` | Split the feature dataset (see below) and use the latest data split for training and test on historical split of the data. This allows the model to be trained up to the most recent data point, while avoiding overfitting. However, you should be careful to understand the unorthodox nature of this parameter before employing it. <br> **Datatype:** Boolean. <br> Default: `False` (no reversal).
| `shuffle_after_split` | Split the data into train and test sets, and then shuffle both sets individually. <br> **Datatype:** Boolean. <br> Default: `False`.
| `buffer_train_data_candles` | Cut `buffer_train_data_candles` off the beginning and end of the training data *after* the indicators were populated. The main example use is when predicting maxima and minima, the argrelextrema function cannot know the maxima/minima at the edges of the timerange. To improve model accuracy, it is best to compute argrelextrema on the full timerange and then use this function to cut off the edges (buffer) by the kernel. In another case, if the targets are set to a shifted price movement, this buffer is unnecessary because the shifted candles at the end of the timerange will be NaN and FreqAI will automatically cut those off of the training dataset.<br> **Datatype:** Boolean. <br> Default: `False`.
| `buffer_train_data_candles` | Cut `buffer_train_data_candles` off the beginning and end of the training data *after* the indicators were populated. The main example use is when predicting maxima and minima, the argrelextrema function cannot know the maxima/minima at the edges of the timerange. To improve model accuracy, it is best to compute argrelextrema on the full timerange and then use this function to cut off the edges (buffer) by the kernel. In another case, if the targets are set to a shifted price movement, this buffer is unnecessary because the shifted candles at the end of the timerange will be NaN and FreqAI will automatically cut those off of the training dataset.<br> **Datatype:** Integer. <br> Default: `0`.
### Data split parameters
@@ -84,6 +84,7 @@ Mandatory parameters are marked as **Required** and have to be set in one of the
| `add_state_info` | Tell FreqAI to include state information in the feature set for training and inferencing. The current state variables include trade duration, current profit, trade position. This is only available in dry/live runs, and is automatically switched to false for backtesting. <br> **Datatype:** bool. <br> Default: `False`.
| `net_arch` | Network architecture which is well described in [`stable_baselines3` doc](https://stable-baselines3.readthedocs.io/en/master/guide/custom_policy.html#examples). In summary: `[<shared layers>, dict(vf=[<non-shared value network layers>], pi=[<non-shared policy network layers>])]`. By default this is set to `[128, 128]`, which defines 2 shared hidden layers with 128 units each.
| `randomize_starting_position` | Randomize the starting point of each episode to avoid overfitting. <br> **Datatype:** bool. <br> Default: `False`.
| `drop_ohlc_from_features` | Do not include the normalized ohlc data in the feature set passed to the agent during training (ohlc will still be used for driving the environment in all cases) <br> **Datatype:** Boolean. <br> **Default:** `False`
### PyTorch parameters

10
docs/freqai-reinforcement-learning.md
View File

@@ -55,7 +55,7 @@ where `ReinforcementLearner` will use the templated `ReinforcementLearner` from
dataframe["&-action"] = 0
```
Most of the function remains the same as for typical Regressors, however, the function above shows how the strategy must pass the raw price data to the agent so that it has access to raw OHLCV in the training environment:
Most of the function remains the same as for typical Regressors, however, the function below shows how the strategy must pass the raw price data to the agent so that it has access to raw OHLCV in the training environment:
```python
def feature_engineering_standard(self, dataframe, **kwargs):
@@ -176,9 +176,11 @@ As you begin to modify the strategy and the prediction model, you will quickly r
factor = 100
pair = self.pair.replace(':', '')
# you can use feature values from dataframe
# Assumes the shifted RSI indicator has been generated in the strategy.
rsi_now = self.raw_features[f"%-rsi-period-10_shift-1_{self.pair}_"
rsi_now = self.raw_features[f"%-rsi-period_10_shift-1_{pair}_"
f"{self.config['timeframe']}"].iloc[self._current_tick]
# reward agent for entering trades
@@ -246,13 +248,13 @@ FreqAI also provides a built in episodic summary logger called `self.tensorboard
"""
def calculate_reward(self, action: int) -> float:
if not self._is_valid(action):
self.tensorboard_log("is_valid")
self.tensorboard_log("invalid")
return -2
```
!!! Note
The `self.tensorboard_log()` function is designed for tracking incremented objects only i.e. events, actions inside the training environment. If the event of interest is a float, the float can be passed as the second argument e.g. `self.tensorboard_log("float_metric1", 0.23)` would add 0.23 to `float_metric`. In this case you can also disable incrementing using `inc=False` parameter.
The `self.tensorboard_log()` function is designed for tracking incremented objects only i.e. events, actions inside the training environment. If the event of interest is a float, the float can be passed as the second argument e.g. `self.tensorboard_log("float_metric1", 0.23)`. In this case the metric values are not incremented.
### Choosing a base environment

3
docs/freqai-running.md
View File

@@ -128,6 +128,9 @@ The FreqAI specific parameter `label_period_candles` defines the offset (number
You can choose to adopt a continual learning scheme by setting `"continual_learning": true` in the config. By enabling `continual_learning`, after training an initial model from scratch, subsequent trainings will start from the final model state of the preceding training. This gives the new model a "memory" of the previous state. By default, this is set to `False` which means that all new models are trained from scratch, without input from previous models.
???+ danger "Continual learning enforces a constant parameter space"
Since `continual_learning` means that the model parameter space *cannot* change between trainings, `principal_component_analysis` is automatically disabled when `continual_learning` is enabled. Hint: PCA changes the parameter space and the number of features, learn more about PCA [here](freqai-feature-engineering.md#data-dimensionality-reduction-with-principal-component-analysis).
## Hyperopt
You can hyperopt using the same command as for [typical Freqtrade hyperopt](hyperopt.md):

2
docs/includes/protections.md
View File

@@ -149,7 +149,7 @@ The below example assumes a timeframe of 1 hour:
* Locks each pair after selling for an additional 5 candles (`CooldownPeriod`), giving other pairs a chance to get filled.
* Stops trading for 4 hours (`4 * 1h candles`) if the last 2 days (`48 * 1h candles`) had 20 trades, which caused a max-drawdown of more than 20%. (`MaxDrawdown`).
* Stops trading if more than 4 stoploss occur for all pairs within a 1 day (`24 * 1h candles`) limit (`StoplossGuard`).
* Locks all pairs that had 4 Trades within the last 6 hours (`6 * 1h candles`) with a combined profit ratio of below 0.02 (<2%) (`LowProfitPairs`).
* Locks all pairs that had 2 Trades within the last 6 hours (`6 * 1h candles`) with a combined profit ratio of below 0.02 (<2%) (`LowProfitPairs`).
* Locks all pairs for 2 candles that had a profit of below 0.01 (<1%) within the last 24h (`24 * 1h candles`), a minimum of 4 trades.
``` python

6
docs/producer-consumer.md
View File

@@ -42,14 +42,14 @@ Enable subscribing to an instance by adding the `external_message_consumer` sect
| `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.port` | **Required.** The port matching the above host.<br>*Defaults to `8080`.*<br> **Datatype:** Integer
| `producers.secure` | **Optional.** Use ssl in websockets connection. Default False.<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.
| `ping_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.
| `remove_entry_exit_signals` | Remove signal columns from the dataframe (set them to 0) on dataframe receipt.<br>*Defaults to `False`.*<br> **Datatype:** Boolean.
| `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.

4
docs/requirements-docs.txt
View File

@@ -1,6 +1,6 @@
markdown==3.3.7
mkdocs==1.4.2
mkdocs-material==9.0.15
mkdocs-material==9.1.5
mdx_truly_sane_lists==1.3
pymdown-extensions==9.9.2
pymdown-extensions==9.10
jinja2==3.1.2

12
docs/strategy-callbacks.md
View File

@@ -51,7 +51,8 @@ During hyperopt, this runs only once at startup.
## Bot loop start
A simple callback which is called once at the start of every bot throttling iteration (roughly every 5 seconds, unless configured differently).
A simple callback which is called once at the start of every bot throttling iteration in dry/live mode (roughly every 5
seconds, unless configured differently) or once per candle in backtest/hyperopt mode.
This can be used to perform calculations which are pair independent (apply to all pairs), loading of external data, etc.
``` python
@@ -61,11 +62,12 @@ class AwesomeStrategy(IStrategy):
# ... populate_* methods
def bot_loop_start(self, **kwargs) -> None:
def bot_loop_start(self, current_time: datetime, **kwargs) -> None:
"""
Called at the start of the bot iteration (one loop).
Might be used to perform pair-independent tasks
(e.g. gather some remote resource for comparison)
:param current_time: datetime object, containing the current datetime
:param **kwargs: Ensure to keep this here so updates to this won't break your strategy.
"""
if self.config['runmode'].value in ('live', 'dry_run'):
@@ -316,11 +318,11 @@ class AwesomeStrategy(IStrategy):
# evaluate highest to lowest, so that highest possible stop is used
if current_profit > 0.40:
return stoploss_from_open(0.25, current_profit, is_short=trade.is_short)
return stoploss_from_open(0.25, current_profit, is_short=trade.is_short, leverage=trade.leverage)
elif current_profit > 0.25:
return stoploss_from_open(0.15, current_profit, is_short=trade.is_short)
return stoploss_from_open(0.15, current_profit, is_short=trade.is_short, leverage=trade.leverage)
elif current_profit > 0.20:
return stoploss_from_open(0.07, current_profit, is_short=trade.is_short)
return stoploss_from_open(0.07, current_profit, is_short=trade.is_short, leverage=trade.leverage)
# return maximum stoploss value, keeping current stoploss price unchanged
return 1

15
docs/strategy-customization.md
View File

@@ -881,7 +881,7 @@ All columns of the informative dataframe will be available on the returning data
### *stoploss_from_open()*
Stoploss values returned from `custom_stoploss` must specify a percentage relative to `current_rate`, but sometimes you may want to specify a stoploss relative to the open price instead. `stoploss_from_open()` is a helper function to calculate a stoploss value that can be returned from `custom_stoploss` which will be equivalent to the desired percentage above the open price.
Stoploss values returned from `custom_stoploss` must specify a percentage relative to `current_rate`, but sometimes you may want to specify a stoploss relative to the entry point instead. `stoploss_from_open()` is a helper function to calculate a stoploss value that can be returned from `custom_stoploss` which will be equivalent to the desired trade profit above the entry point.
??? Example "Returning a stoploss relative to the open price from the custom stoploss function"
@@ -889,6 +889,8 @@ Stoploss values returned from `custom_stoploss` must specify a percentage relati
If we want a stop price at 7% above the open price we can call `stoploss_from_open(0.07, current_profit, False)` which will return `0.1157024793`. 11.57% below $121 is $107, which is the same as 7% above $100.
This function will consider leverage - so at 10x leverage, the actual stoploss would be 0.7% above $100 (0.7% * 10x = 7%).
``` python
@@ -907,7 +909,7 @@ Stoploss values returned from `custom_stoploss` must specify a percentage relati
# once the profit has risen above 10%, keep the stoploss at 7% above the open price
if current_profit > 0.10:
return stoploss_from_open(0.07, current_profit, is_short=trade.is_short)
return stoploss_from_open(0.07, current_profit, is_short=trade.is_short, leverage=trade.leverage)
return 1
@@ -1038,11 +1040,10 @@ from datetime import timedelta, datetime, timezone
# Within populate indicators (or populate_buy):
if self.config['runmode'].value in ('live', 'dry_run'):
# fetch closed trades for the last 2 days
trades = Trade.get_trades([Trade.pair == metadata['pair'],
Trade.open_date > datetime.utcnow() - timedelta(days=2),
Trade.is_open.is_(False),
]).all()
# fetch closed trades for the last 2 days
trades = Trade.get_trades_proxy(
pair=metadata['pair'], is_open=False,
open_date=datetime.now(timezone.utc) - timedelta(days=2))
# Analyze the conditions you'd like to lock the pair .... will probably be different for every strategy
sumprofit = sum(trade.close_profit for trade in trades)
if sumprofit < 0:

44
docs/utils.md
View File

@@ -955,3 +955,47 @@ Print trades with id 2 and 3 as json
``` bash
freqtrade show-trades --db-url sqlite:///tradesv3.sqlite --trade-ids 2 3 --print-json
```
### Strategy-Updater
Updates listed strategies or all strategies within the strategies folder to be v3 compliant.
If the command runs without --strategy-list then all strategies inside the strategies folder will be converted.
Your original strategy will remain available in the `user_data/strategies_orig_updater/` directory.
!!! Warning "Conversion results"
Strategy updater will work on a "best effort" approach. Please do your due diligence and verify the results of the conversion.
We also recommend to run a python formatter (e.g. `black`) to format results in a sane manner.
```
usage: freqtrade strategy-updater [-h] [-v] [--logfile FILE] [-V] [-c PATH]
[-d PATH] [--userdir PATH]
[--strategy-list STRATEGY_LIST [STRATEGY_LIST ...]]
options:
-h, --help show this help message and exit
--strategy-list STRATEGY_LIST [STRATEGY_LIST ...]
Provide a space-separated list of strategies to
backtest. Please note that timeframe needs to be set
either in config or via command line. When using this
together with `--export trades`, the strategy-name is
injected into the filename (so `backtest-data.json`
becomes `backtest-data-SampleStrategy.json`
Common arguments:
-v, --verbose Verbose mode (-vv for more, -vvv to get all messages).
--logfile FILE, --log-file FILE
Log to the file specified. Special values are:
'syslog', 'journald'. See the documentation for more
details.
-V, --version show program's version number and exit
-c PATH, --config PATH
Specify configuration file (default:
`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, --data-dir PATH
Path to directory with historical backtesting data.
--userdir PATH, --user-data-dir PATH
Path to userdata directory.
```