feature: introducing pricing predictors (pricers)

experiments/procesing/pricers/base.py

@@ -0,0 +1,28 @@
+from abc import ABC, abstractmethod
+import numpy as np
+import pandas as pd
+
+
+class PricingFunction(ABC):
+    """
+    Abstract base for pricing functions.
+    Defines the mapping f: StateSpace -> prices
+    """
+
+    @abstractmethod
+    def fit(self, historical_data: pd.DataFrame):
+        """Train/calibrate the pricing function on historical data"""
+        pass
+
+    @abstractmethod
+    def predict(self, state_space) -> np.ndarray:
+        """
+        Generate prices given current state space.
+
+        Args:
+            state_space: StateSpace object containing demand, prices, session features
+
+        Returns:
+            prices: price vector P_{t+1} in R^n
+        """
+        pass