stable/freqtrade/data/history.py

340 lines
13 KiB
Python
Raw Normal View History

2018-12-16 08:58:46 +00:00
"""
Handle historic data (ohlcv).
Includes:
2018-12-16 08:58:46 +00:00
* load data for a pair (or a list of pairs) from disk
* download data from exchange and store to disk
"""
import logging
import operator
from datetime import datetime
2018-12-15 12:54:35 +00:00
from pathlib import Path
from typing import Any, Dict, List, Optional, Tuple
import arrow
2018-12-15 13:28:37 +00:00
from pandas import DataFrame
from freqtrade import OperationalException, misc
from freqtrade.configuration import TimeRange
2018-12-15 13:28:37 +00:00
from freqtrade.data.converter import parse_ticker_dataframe
from freqtrade.exchange import Exchange, timeframe_to_minutes
logger = logging.getLogger(__name__)
def trim_tickerlist(tickerlist: List[Dict], timerange: TimeRange) -> List[Dict]:
"""
Trim tickerlist based on given timerange
"""
if not tickerlist:
return tickerlist
start_index = 0
stop_index = len(tickerlist)
if timerange.starttype == 'line':
stop_index = timerange.startts
if timerange.starttype == 'index':
start_index = timerange.startts
elif timerange.starttype == 'date':
while (start_index < len(tickerlist) and
tickerlist[start_index][0] < timerange.startts * 1000):
start_index += 1
if timerange.stoptype == 'line':
2019-08-15 04:35:50 +00:00
start_index = max(len(tickerlist) + timerange.stopts, 0)
if timerange.stoptype == 'index':
stop_index = timerange.stopts
elif timerange.stoptype == 'date':
while (stop_index > 0 and
tickerlist[stop_index-1][0] > timerange.stopts * 1000):
stop_index -= 1
if start_index > stop_index:
raise ValueError(f'The timerange [{timerange.startts},{timerange.stopts}] is incorrect')
return tickerlist[start_index:stop_index]
def load_tickerdata_file(datadir: Path, pair: str, ticker_interval: str,
timerange: Optional[TimeRange] = None) -> Optional[list]:
"""
Load a pair from file, either .json.gz or .json
2019-09-05 20:00:16 +00:00
:return: tickerlist or None if unsuccessful
"""
2019-05-21 17:49:02 +00:00
filename = pair_data_filename(datadir, pair, ticker_interval)
pairdata = misc.file_load_json(filename)
if not pairdata:
return []
if timerange:
pairdata = trim_tickerlist(pairdata, timerange)
return pairdata
def store_tickerdata_file(datadir: Path, pair: str,
2019-08-14 18:48:42 +00:00
ticker_interval: str, data: list, is_zip: bool = False):
"""
Stores tickerdata to file
"""
filename = pair_data_filename(datadir, pair, ticker_interval)
misc.file_dump_json(filename, data, is_zip=is_zip)
def load_pair_history(pair: str,
ticker_interval: str,
datadir: Path,
2018-12-16 09:17:11 +00:00
timerange: TimeRange = TimeRange(None, None, 0, 0),
refresh_pairs: bool = False,
exchange: Optional[Exchange] = None,
2019-06-09 12:40:45 +00:00
fill_up_missing: bool = True,
drop_incomplete: bool = True
2018-12-16 09:17:11 +00:00
) -> DataFrame:
"""
Loads cached ticker history for the given pair.
2019-06-09 12:40:45 +00:00
:param pair: Pair to load data for
:param ticker_interval: Ticker-interval (e.g. "5m")
:param datadir: Path to the data storage location.
:param timerange: Limit data to be loaded to this timerange
:param refresh_pairs: Refresh pairs from exchange.
(Note: Requires exchange to be passed as well.)
:param exchange: Exchange object (needed when using "refresh_pairs")
:param fill_up_missing: Fill missing values with "No action"-candles
:param drop_incomplete: Drop last candle assuming it may be incomplete.
2018-12-16 13:14:17 +00:00
:return: DataFrame with ohlcv data
"""
2019-05-17 16:05:36 +00:00
# The user forced the refresh of pairs
2018-12-16 09:17:11 +00:00
if refresh_pairs:
download_pair_history(datadir=datadir,
exchange=exchange,
pair=pair,
ticker_interval=ticker_interval,
timerange=timerange)
2018-12-16 09:17:11 +00:00
2019-01-01 12:42:30 +00:00
pairdata = load_tickerdata_file(datadir, pair, ticker_interval, timerange=timerange)
if pairdata:
if timerange.starttype == 'date' and pairdata[0][0] > timerange.startts * 1000:
logger.warning('Missing data at start for pair %s, data starts at %s',
pair, arrow.get(pairdata[0][0] // 1000).strftime('%Y-%m-%d %H:%M:%S'))
if timerange.stoptype == 'date' and pairdata[-1][0] < timerange.stopts * 1000:
logger.warning('Missing data at end for pair %s, data ends at %s',
pair,
arrow.get(pairdata[-1][0] // 1000).strftime('%Y-%m-%d %H:%M:%S'))
return parse_ticker_dataframe(pairdata, ticker_interval, pair=pair,
2019-06-09 12:40:45 +00:00
fill_missing=fill_up_missing,
drop_incomplete=drop_incomplete)
else:
2019-05-17 16:05:36 +00:00
logger.warning(
f'No history data for pair: "{pair}", interval: {ticker_interval}. '
'Use --refresh-pairs-cached option or `freqtrade download-data` '
2019-05-17 16:05:36 +00:00
'script to download the data'
)
return None
def load_data(datadir: Path,
ticker_interval: str,
pairs: List[str],
2018-12-16 09:17:11 +00:00
refresh_pairs: bool = False,
exchange: Optional[Exchange] = None,
timerange: TimeRange = TimeRange(None, None, 0, 0),
fill_up_missing: bool = True,
live: bool = False
) -> Dict[str, DataFrame]:
"""
2018-12-16 09:17:11 +00:00
Loads ticker history data for a list of pairs the given parameters
2018-12-16 13:14:17 +00:00
:return: dict(<pair>:<tickerlist>)
"""
result: Dict[str, DataFrame] = {}
if live:
if exchange:
logger.info('Live: Downloading data for all defined pairs ...')
exchange.refresh_latest_ohlcv([(pair, ticker_interval) for pair in pairs])
result = {key[0]: value for key, value in exchange._klines.items() if value is not None}
else:
raise OperationalException(
"Exchange needs to be initialized when using live data."
)
else:
logger.info('Using local backtesting data ...')
for pair in pairs:
hist = load_pair_history(pair=pair, ticker_interval=ticker_interval,
datadir=datadir, timerange=timerange,
refresh_pairs=refresh_pairs,
exchange=exchange,
fill_up_missing=fill_up_missing)
if hist is not None:
result[pair] = hist
return result
def pair_data_filename(datadir: Path, pair: str, ticker_interval: str) -> Path:
2019-05-21 17:49:02 +00:00
pair_s = pair.replace("/", "_")
filename = datadir.joinpath(f'{pair_s}-{ticker_interval}.json')
2019-05-21 17:49:02 +00:00
return filename
def load_cached_data_for_updating(datadir: Path, pair: str, ticker_interval: str,
timerange: Optional[TimeRange]) -> Tuple[List[Any],
Optional[int]]:
"""
Load cached data to download more data.
If timerange is passed in, checks wether data from an before the stored data will be downloaded.
If that's the case than what's available should be completely overwritten.
Only used by download_pair_history().
"""
since_ms = None
# user sets timerange, so find the start time
if timerange:
if timerange.starttype == 'date':
since_ms = timerange.startts * 1000
elif timerange.stoptype == 'line':
num_minutes = timerange.stopts * timeframe_to_minutes(ticker_interval)
since_ms = arrow.utcnow().shift(minutes=num_minutes).timestamp * 1000
# read the cached file
# Intentionally don't pass timerange in - since we need to load the full dataset.
data = load_tickerdata_file(datadir, pair, ticker_interval)
2019-08-14 18:48:42 +00:00
# remove the last item, could be incomplete candle
if data:
data.pop()
else:
data = []
if data:
if since_ms and since_ms < data[0][0]:
# Earlier data than existing data requested, redownload all
data = []
else:
# a part of the data was already downloaded, so download unexist data only
since_ms = data[-1][0] + 1
return (data, since_ms)
def download_pair_history(datadir: Path,
2019-05-17 16:05:36 +00:00
exchange: Optional[Exchange],
pair: str,
ticker_interval: str = '5m',
timerange: Optional[TimeRange] = None) -> bool:
"""
Download the latest ticker intervals from the exchange for the pair passed in parameters
The data is downloaded starting from the last correct ticker interval data that
exists in a cache. If timerange starts earlier than the data in the cache,
the full data will be redownloaded
Based on @Rybolov work: https://github.com/rybolov/freqtrade-data
2019-05-17 16:05:36 +00:00
:param pair: pair to download
:param ticker_interval: ticker interval
:param timerange: range of time to download
2018-12-16 13:14:17 +00:00
:return: bool with success state
"""
2019-05-17 16:05:36 +00:00
if not exchange:
raise OperationalException(
"Exchange needs to be initialized when downloading pair history data"
)
2018-12-16 09:29:53 +00:00
try:
2019-05-17 16:05:36 +00:00
logger.info(
f'Download history data for pair: "{pair}", interval: {ticker_interval} '
f'and store in {datadir}.'
)
2019-08-14 18:48:42 +00:00
data, since_ms = load_cached_data_for_updating(datadir, pair, ticker_interval, timerange)
2018-12-16 09:29:53 +00:00
logger.debug("Current Start: %s", misc.format_ms_time(data[1][0]) if data else 'None')
logger.debug("Current End: %s", misc.format_ms_time(data[-1][0]) if data else 'None')
2018-12-16 09:29:53 +00:00
# Default since_ms to 30 days if nothing is given
new_data = exchange.get_historic_ohlcv(pair=pair, ticker_interval=ticker_interval,
since_ms=since_ms if since_ms
else
int(arrow.utcnow().shift(
days=-30).float_timestamp) * 1000)
2018-12-16 09:29:53 +00:00
data.extend(new_data)
2018-12-16 09:29:53 +00:00
logger.debug("New Start: %s", misc.format_ms_time(data[0][0]))
logger.debug("New End: %s", misc.format_ms_time(data[-1][0]))
2019-08-14 18:48:42 +00:00
store_tickerdata_file(datadir, pair, ticker_interval, data=data)
2018-12-16 09:29:53 +00:00
return True
2019-05-17 16:05:36 +00:00
2019-05-21 17:49:02 +00:00
except Exception as e:
2019-05-17 16:05:36 +00:00
logger.error(
2019-05-21 17:49:02 +00:00
f'Failed to download history data for pair: "{pair}", interval: {ticker_interval}. '
f'Error: {e}'
2019-05-17 16:05:36 +00:00
)
2019-01-31 05:51:03 +00:00
return False
def refresh_backtest_ohlcv_data(exchange: Exchange, pairs: List[str], timeframes: List[str],
dl_path: Path, timerange: TimeRange,
erase=False) -> List[str]:
"""
Refresh stored ohlcv data for backtesting and hyperopt operations.
Used by freqtrade download-data
:return: Pairs not available
"""
pairs_not_available = []
for pair in pairs:
if pair not in exchange.markets:
pairs_not_available.append(pair)
logger.info(f"Skipping pair {pair}...")
continue
for ticker_interval in timeframes:
dl_file = pair_data_filename(dl_path, pair, ticker_interval)
if erase and dl_file.exists():
logger.info(
f'Deleting existing data for pair {pair}, interval {ticker_interval}.')
dl_file.unlink()
logger.info(f'Downloading pair {pair}, interval {ticker_interval}.')
download_pair_history(datadir=dl_path, exchange=exchange,
pair=pair, ticker_interval=str(ticker_interval),
timerange=timerange)
return pairs_not_available
def get_timeframe(data: Dict[str, DataFrame]) -> Tuple[arrow.Arrow, arrow.Arrow]:
"""
Get the maximum timeframe for the given backtest data
:param data: dictionary with preprocessed backtesting data
:return: tuple containing min_date, max_date
"""
timeframe = [
(arrow.get(frame['date'].min()), arrow.get(frame['date'].max()))
for frame in data.values()
]
return min(timeframe, key=operator.itemgetter(0))[0], \
max(timeframe, key=operator.itemgetter(1))[1]
def validate_backtest_data(data: DataFrame, pair: str, min_date: datetime,
max_date: datetime, ticker_interval_mins: int) -> bool:
"""
Validates preprocessed backtesting data for missing values and shows warnings about it that.
:param data: preprocessed backtesting data (as DataFrame)
:param pair: pair used for log output.
:param min_date: start-date of the data
:param max_date: end-date of the data
:param ticker_interval_mins: ticker interval in minutes
"""
# total difference in minutes / interval-minutes
expected_frames = int((max_date - min_date).total_seconds() // 60 // ticker_interval_mins)
found_missing = False
dflen = len(data)
if dflen < expected_frames:
found_missing = True
logger.warning("%s has missing frames: expected %s, got %s, that's %s missing values",
pair, expected_frames, dflen, expected_frames - dflen)
return found_missing