stable/scripts/convert_backtestdata.py

203 lines
6.6 KiB
Python
Raw Normal View History

2018-03-26 18:18:14 +00:00
#!/usr/bin/env python3
"""
Script to display when the bot will buy a specific pair
Mandatory Cli parameters:
-p / --pair: pair to examine
Optional Cli parameters
-d / --datadir: path to pair backtest data
--timerange: specify what timerange of data to use.
-l / --live: Live, to download the latest ticker for the pair
"""
import sys
from argparse import Namespace
from os import path
import glob
import json
import re
from typing import List, Dict
2018-03-30 21:30:23 +00:00
import gzip
2018-03-26 18:18:14 +00:00
from freqtrade.arguments import Arguments
from freqtrade import misc
from freqtrade.logger import Logger
from pandas import DataFrame
from freqtrade.constants import Constants
2018-03-26 18:18:14 +00:00
import dateutil.parser
2018-03-31 15:30:11 +00:00
logger = Logger(name="freqtrade").get_logger()
2018-03-26 18:18:14 +00:00
2018-03-30 21:30:23 +00:00
def load_old_file(filename) -> (List[Dict], bool):
2018-03-26 18:18:14 +00:00
if not path.isfile(filename):
logger.warning("filename %s does not exist", filename)
2018-03-30 21:30:23 +00:00
return (None, False)
2018-03-26 18:18:14 +00:00
logger.debug('Loading ticker data from file %s', filename)
pairdata = None
2018-03-30 21:30:23 +00:00
if filename.endswith('.gz'):
logger.debug('Loading ticker data from file %s', filename)
is_zip = True
with gzip.open(filename) as tickerdata:
pairdata = json.load(tickerdata)
else:
is_zip = False
with open(filename) as tickerdata:
pairdata = json.load(tickerdata)
return (pairdata, is_zip)
2018-03-26 18:18:14 +00:00
def parse_old_backtest_data(ticker) -> DataFrame:
"""
Reads old backtest data
Format: "O": 8.794e-05,
"H": 8.948e-05,
"L": 8.794e-05,
"C": 8.88e-05,
"V": 991.09056638,
"T": "2017-11-26T08:50:00",
"BV": 0.0877869
"""
columns = {'C': 'close', 'V': 'volume', 'O': 'open',
'H': 'high', 'L': 'low', 'T': 'date'}
frame = DataFrame(ticker) \
.rename(columns=columns)
if 'BV' in frame:
frame.drop('BV', 1, inplace=True)
if not 'date' in frame:
logger.warning("Date not in frame - probably not a Ticker file")
return None
2018-03-26 18:18:14 +00:00
frame.sort_values('date', inplace=True)
return frame
def convert_dataframe(frame: DataFrame):
"""Convert dataframe to new format"""
# reorder columns:
cols = ['date', 'open', 'high', 'low', 'close', 'volume']
frame = frame[cols]
# Make sure parsing/printing data is assumed to be UTC
2018-03-26 18:18:14 +00:00
frame['date'] = frame['date'].apply(
lambda d: int(dateutil.parser.parse(d+'+00:00').timestamp()) * 1000)
frame['date'] = frame['date'].astype('int64')
2018-03-26 18:18:14 +00:00
# Convert columns one by one to preserve type.
by_column = [frame[x].values.tolist() for x in frame.columns]
return list(list(x) for x in zip(*by_column))
2018-03-31 15:29:52 +00:00
def convert_file(filename: str, filename_new: str) -> None:
2018-03-26 21:16:41 +00:00
"""Converts a file from old format to ccxt format"""
2018-03-30 21:30:23 +00:00
(pairdata, is_zip) = load_old_file(filename)
2018-03-31 15:29:52 +00:00
if pairdata and type(pairdata) is list:
2018-03-26 18:18:14 +00:00
if type(pairdata[0]) is list:
logger.error("pairdata for %s already in new format", filename)
2018-03-31 15:29:52 +00:00
return
2018-03-26 18:18:14 +00:00
frame = parse_old_backtest_data(pairdata)
# Convert frame to new format
if frame is not None:
frame1 = convert_dataframe(frame)
misc.file_dump_json(filename_new, frame1, is_zip)
2018-03-26 18:18:14 +00:00
def convert_main(args: Namespace) -> None:
"""
converts a folder given in --datadir from old to new format to support ccxt
"""
2018-03-31 15:24:25 +00:00
workdir = path.join(args.datadir, "")
2018-03-30 21:30:23 +00:00
logger.info("Workdir: %s", workdir)
2018-03-26 18:18:14 +00:00
for filename in glob.glob(workdir + "*.json"):
# swap currency names
ret = re.search(r'[A-Z_]{7,}', path.basename(filename))
if args.norename:
filename_new = filename
else:
if not ret:
logger.warning("file %s could not be converted, could not extract currencies",
filename)
continue
pair = ret.group(0)
currencies = pair.split("_")
if len(currencies) != 2:
logger.warning("file %s could not be converted, could not extract currencies",
filename)
continue
ret_integer = re.search(r'\d+(?=\.json)', path.basename(filename))
ret_string = re.search(r'(\d+[mhdw])(?=\.json)', path.basename(filename))
if ret_integer:
minutes = int(ret_integer.group(0))
interval = str(minutes) + 'm' # default to adding 'm' to end of minutes for new interval name
# but check if there is a mapping between int and string also
for str_interval, minutes_interval in Constants.TICKER_INTERVAL_MINUTES.items():
if minutes_interval == minutes:
interval = str_interval
break
# change order on pairs if old ticker interval found
filename_new = path.join(path.dirname(filename),
"{}_{}-{}.json".format(currencies[1],
currencies[0], interval))
elif ret_string:
interval = ret_string.group(0)
filename_new = path.join(path.dirname(filename),
"{}_{}-{}.json".format(currencies[0],
currencies[1], interval))
else:
2018-03-26 18:18:14 +00:00
logger.warning("file %s could not be converted, interval not found", filename)
continue
2018-03-31 15:24:25 +00:00
2018-03-26 18:18:14 +00:00
logger.debug("Converting and renaming %s to %s", filename, filename_new)
convert_file(filename, filename_new)
def convert_parse_args(args: List[str]) -> Namespace:
"""
Parse args passed to the script
:param args: Cli arguments
:return: args: Array with all arguments
"""
arguments = Arguments(args, 'Convert datafiles')
arguments.parser.add_argument(
'-d', '--datadir',
help='path to backtest data (default: %(default)s',
dest='datadir',
default=path.join('freqtrade', 'tests', 'testdata'),
type=str,
metavar='PATH',
)
arguments.parser.add_argument(
'-n', '--norename',
help='don''t rename files from BTC_<PAIR> to <PAIR>_BTC - '
'Note that not renaming will overwrite source files',
dest='norename',
default=False,
action='store_true'
)
return arguments.parse_args()
def main(sysargv: List[str]) -> None:
"""
This function will initiate the bot and start the trading loop.
:return: None
"""
logger.info('Starting Dataframe conversation')
convert_main(convert_parse_args(sysargv))
if __name__ == '__main__':
main(sys.argv[1:])