Increase performance by only predicting on most recent candle instead of full strat provided dataframe. Collect predictions and store them so that we can feed true predictions back to strategy (so that frequi isnt updating historic predictions based on newly trained models).
This commit is contained in:
parent
2f1a2c1cd7
commit
a20651efd8
@ -1,4 +1,5 @@
|
||||
|
||||
import copy
|
||||
import json
|
||||
import logging
|
||||
from pathlib import Path
|
||||
@ -24,6 +25,7 @@ class FreqaiDataDrawer:
|
||||
self.pair_dict: Dict[str, Any] = {}
|
||||
# dictionary holding all actively inferenced models in memory given a model filename
|
||||
self.model_dictionary: Dict[str, Any] = {}
|
||||
self.model_return_values: Dict[str, Any] = {}
|
||||
self.pair_data_dict: Dict[str, Any] = {}
|
||||
self.full_path = full_path
|
||||
self.load_drawer_from_disk()
|
||||
@ -83,3 +85,27 @@ class FreqaiDataDrawer:
|
||||
self.pair_dict[p]['priority'] -= 1
|
||||
# send pair to end of queue
|
||||
self.pair_dict[pair]['priority'] = len(self.pair_dict)
|
||||
|
||||
def set_initial_return_values(self, pair, dh):
|
||||
self.model_return_values[pair] = {}
|
||||
self.model_return_values[pair]['predictions'] = dh.full_predictions
|
||||
self.model_return_values[pair]['do_preds'] = dh.full_do_predict
|
||||
self.model_return_values[pair]['target_mean'] = dh.full_target_mean
|
||||
self.model_return_values[pair]['target_std'] = dh.full_target_std
|
||||
|
||||
def append_model_predictions(self, pair, predictions, do_preds,
|
||||
target_mean, target_std, dh) -> None:
|
||||
|
||||
pred_store = self.model_return_values[pair]['predictions']
|
||||
do_pred_store = self.model_return_values[pair]['do_preds']
|
||||
tm_store = self.model_return_values[pair]['target_mean']
|
||||
ts_store = self.model_return_values[pair]['target_std']
|
||||
pred_store = np.append(pred_store[1:], predictions[-1])
|
||||
do_pred_store = np.append(do_pred_store[1:], do_preds[-1])
|
||||
tm_store = np.append(tm_store[1:], target_mean)
|
||||
ts_store = np.append(ts_store[1:], target_std)
|
||||
|
||||
dh.full_predictions = copy.deepcopy(pred_store)
|
||||
dh.full_do_predict = copy.deepcopy(do_pred_store)
|
||||
dh.full_target_mean = copy.deepcopy(tm_store)
|
||||
dh.full_target_std = copy.deepcopy(ts_store)
|
||||
|
@ -623,7 +623,7 @@ class FreqaiDataKitchen:
|
||||
Append backtest prediction from current backtest period to all previous periods
|
||||
"""
|
||||
|
||||
ones = np.ones(len_dataframe)
|
||||
ones = np.ones(len(predictions))
|
||||
target_mean, target_std = ones * self.data["target_mean"], ones * self.data["target_std"]
|
||||
|
||||
self.full_predictions = np.append(self.full_predictions, predictions)
|
||||
|
@ -219,8 +219,16 @@ class IFreqaiModel(ABC):
|
||||
|
||||
self.check_if_feature_list_matches_strategy(dataframe, dh)
|
||||
|
||||
preds, do_preds = self.predict(dataframe, dh)
|
||||
dh.append_predictions(preds, do_preds, len(dataframe))
|
||||
if metadata['pair'] not in self.data_drawer.model_return_values:
|
||||
preds, do_preds = self.predict(dataframe, dh)
|
||||
dh.append_predictions(preds, do_preds, len(dataframe))
|
||||
dh.fill_predictions(len(dataframe))
|
||||
self.data_drawer.set_initial_return_values(metadata['pair'], dh)
|
||||
else:
|
||||
preds, do_preds = self.predict(dataframe.iloc[-2:], dh)
|
||||
self.data_drawer.append_model_predictions(metadata['pair'], preds, do_preds,
|
||||
self.dh.data["target_mean"],
|
||||
self.dh.data["target_std"], dh)
|
||||
|
||||
return dh
|
||||
|
||||
|
Loading…
Reference in New Issue
Block a user