currently debugging mismatched shape error when computing weights and

app.py

@@ -1,5 +1,4 @@
 from flask import Flask, request, jsonify, Response
-
 from nn.nn import NN
 from nn import train as train_nn
 from nn import activation

nn/activation.py

@@ -26,7 +26,10 @@ def relu(x):
 
 
 def relu_prime(x):
-    return
+    if x > 0:
+        return 1
+    else:
+        return 0
 
 
 def sigmoid(x):
@@ -43,4 +46,4 @@ def tanh(x):
 
 
 def tanh_prime(x):
-    return
+    return 1 - np.tanh(x)**2

nn/nn.py

@@ -1,6 +1,5 @@
 from typing import Callable
 import pandas as pd
-import numpy as np
 
 
 class NN:
@@ -24,10 +23,7 @@ class NN:
         self.target = target
         self.data = data
 
-        self.wh: np.array = None
-        self.wo: np.array = None
-        self.bh: np.array = None
-        self.bo: np.array = None
+        self.loss_hist: list[float] = None
         self.func_prime: Callable = None
         self.func: Callable = None
         self.df: pd.DataFrame = None
@@ -52,18 +48,6 @@ class NN:
         assert isinstance(f, Callable)
         self.func_prime = f
 
-    def set_bh(self, bh: np.array) -> None:
-        self.bh = bh
-
-    def set_wh(self, wh: np.array) -> None:
-        self.wh = wh
-
-    def set_bo(self, bo: np.array) -> None:
-        self.bo = bo
-
-    def set_wo(self, wo: np.array) -> None:
-        self.wo = wo
-
     @classmethod
     def from_dict(cls, dct):
         """ Creates an instance of NN given a dictionary

nn/train.py

@@ -1,7 +1,6 @@
 from sklearn.model_selection import train_test_split
 from typing import Callable
 from nn.nn import NN
-import pandas as pd
 import numpy as np
 
 
@@ -11,40 +10,50 @@ def init_weights_biases(nn: NN) -> None:
     wh = np.random.randn(nn.input_size, nn.hidden_size) * \
         np.sqrt(2 / nn.input_size)
     wo = np.random.randn(nn.hidden_size, 1) * np.sqrt(2 / nn.hidden_size)
-    nn.set_bh(bh)
-    nn.set_bo(bo)
-    nn.set_wh(wh)
-    nn.set_wo(wo)
+    return wh, wo, bh, bo
 
 
 def train(nn: NN) -> dict:
-    init_weights_biases(nn=nn)
+    wh, wo, bh, bo = init_weights_biases(nn=nn)
     X_train, X_test, y_train, y_test = train_test_split(
         nn.X,
         nn.y,
         test_size=nn.test_size,
     )
 
+    mse: float = 0.0
+    loss_hist: list[float] = []
     for _ in range(nn.epochs):
         # compute hidden output
         hidden_output = compute_node(
             data=X_train.to_numpy(),
-            weights=nn.wh,
-            biases=nn.bh,
+            weights=wh,
+            biases=bh,
             func=nn.func,
         )
 
         # compute output layer
         y_hat = compute_node(
             data=hidden_output,
-            weights=nn.wo,
-            biases=nn.bo,
+            weights=wo,
+            biases=bo,
             func=nn.func,
         )
-
+        # compute error & store it
+        error = y_hat - y_train
         mse = mean_squared_error(y_train, y_hat)
+        loss_hist.append(mse)
 
-    return {"mse": mse}
+        # update weights & biases using gradient descent after
+        # computing derivatives.
+        wh -= (nn.learning_rate * hidden_weight_prime(X_train, error))
+        wo -= (nn.learning_rate * output_weight_prime(hidden_output, error))
+        bh -= (nn.learning_rate * hidden_bias_prime(error))
+        bo -= (nn.learning_rate * output_bias_prime(error))
+    return {
+        "mse": mse,
+        "loss_hist": loss_hist,
+    }
 
 
 def compute_node(data: np.array, weights: np.array, biases: np.array, func: Callable) -> np.array:
@@ -53,3 +62,19 @@ def compute_node(data: np.array, weights: np.array, biases: np.array, func: Call
 
 def mean_squared_error(y: np.array, y_hat: np.array) -> np.array:
     return np.mean((y - y_hat) ** 2)
+
+
+def hidden_weight_prime(data, error):
+    return np.dot(data.T, error)
+
+
+def output_weight_prime(hidden_output, error):
+    return np.dot(hidden_output.T, error)
+
+
+def hidden_bias_prime(error):
+    return np.sum(error, axis=0)
+
+
+def output_bias_prime(error):
+    return np.sum(error, axis=0)

requirements.txt

@@ -1,3 +1,5 @@
-Flask==3.0.0
-pandas==2.1.1
-Requests==2.31.0
+Flask==2.2.3
+numpy==1.25.2
+pandas==1.5.3
+requests==2.28.2
+scikit_learn==1.3.1