48 feature eval computation

src/pysrc/util/feature_eval.py

@@ -0,0 +1,42 @@
+from pysrc.adapters.kraken.historical.historical_data_client import HistoricalDataClient
+from datetime import datetime
+from dateutil.rrule import rrule, DAILY
+from pysrc.util.types import TradeMessage, Asset
+from pysrc.signal.base_feature_generator import BaseFeatureGenerator
+import numpy as np
+
+
+class Evaluator:
+
+    def __init__(self, features: list[str], asset: Asset, start: datetime, end: datetime):
+        self.features = features
+        self.asset = asset
+        self.start = start
+        self.end = end
+        pass
+
+    def _get_data_daterange(self, start: datetime, end: datetime) -> list[TradeMessage]:
+        hist_client = HistoricalDataClient()
+        res = []
+        for date in rrule(DAILY, dtstart=start, until=end):
+            res.extend(hist_client.get_trades(self.coin, date, ""))
+        return res
+
+    def calculate_features(self) -> dict[str, list[float]]:
+        trades = self._get_data_daterange(self.start, self.end)
+        # ^ this returns a list[TradeMessage], but on_tick expects dict[str, TradeMessage]
+        # What is key?
+        snapshots = {}
+        # ^ how to get snapshots?
+        generator_client = BaseFeatureGenerator()
+        feature_dict = generator_client.on_tick(snapshots, trades)
+        return feature_dict[self.asset]
+
+    # Suppose that returns is passed to evaluate_features
+    def evaluate_features(self, target: list[float]):
+        calc_features = self.calculate_features()
+        input_matrix = []
+        for feature in self.features:
+            input_matrix.append(calc_features[feature])
+        input_matrix.append(target)
+        return np.corrocoef(input_matrix)