Eric Meehan
e2034cfa65
Reverted to MyStrategy instead of Cerebro experiments and implemented base functionality for a Matrix logging mechanism. Reviewed-on: #6
54 lines
2.0 KiB
Python
54 lines
2.0 KiB
Python
import os
|
|
import pandas
|
|
|
|
from scipy.stats import linregress
|
|
|
|
from kraken_bot.Model import Model
|
|
|
|
class LinearRegression(Model):
|
|
def __init__(self):
|
|
self.r_value_target = float(os.getenv('R_VALUE_TARGET'))
|
|
self.analysis = pandas.DataFrame({})
|
|
super().__init__(
|
|
interval = 1,
|
|
ascending = False
|
|
)
|
|
|
|
def _linear_regression_for_ohlc_data(self, ohlc_data):
|
|
self.log.debug(f"_linear_regression_for_ohlc_data: ohlc_data=...")
|
|
return linregress(ohlc_data['time'].astype(float), ohlc_data['close'].astype(float))
|
|
|
|
def _filter_asset_pairs(self, tradable_asset_pairs):
|
|
self.log.debug(f"_filter_asset_pairs: tradable_asset_pairs=...")
|
|
return tradable_asset_pairs[
|
|
(tradable_asset_pairs['quote'] == 'ZUSD') &
|
|
(tradable_asset_pairs['status'] == 'online')
|
|
]
|
|
|
|
def _filter_regressions(self, regressions):
|
|
self.log.debug(f"_filter_regressions: regressions=...")
|
|
regressions.columns = ['slope', 'intercept', 'r_value', 'p_value', 'stderr']
|
|
return regressions[
|
|
(regressions.r_value >= self.r_value_target) &
|
|
(regressions.slope > 0)
|
|
]
|
|
|
|
def _order_regressions(self, regressions):
|
|
self.log.debug(f"_order_regressions: regressions=...")
|
|
return regressions.sort_values(by='slope', ascending=False)
|
|
|
|
def since(self):
|
|
return None
|
|
|
|
def update(self, tradable_asset_pairs, ohlc_data):
|
|
self.log.debug(f"update: tradable_asset_pairs=..., ohlc_data=...")
|
|
self.analysis = self._order_regressions(
|
|
self._filter_regressions(
|
|
pandas.DataFrame({
|
|
asset_pair: self._linear_regression_for_ohlc_data(
|
|
ohlc_data[asset_pair][0]
|
|
) for asset_pair in self._filter_asset_pairs(tradable_asset_pairs).index
|
|
}).transpose()
|
|
))
|
|
self.log.debug(self.analysis)
|