after running this code I keep getting the same error:

note:(the data is in excel file (Heights : 16 column) and (Wights:16 column)

I tried to change the epochs_num and it keeps giving the same problem...

import pandas as pd
import matplotlib.pyplot as plt
import numpy as np
# Load the dataset
data = pd.read_csv('heights_weights.csv')
# Plot the data distribution
plt.scatter(data['Height'], data['Weight'], color='b')
plt.xlabel('Height')
plt.ylabel('Weight')
plt.title('Height vs. Weight')
plt.show()
# Define the linear regression model
def linearRegression_model(X, weights):
   y_pred = np.dot(X, weights)
   return y_pred
# Define the update weights function
def linearRegression_update_weights(X, y, weights, learning_rate):
   y_pred = linearRegression_model(X, weights)
   weights_delta = np.dot(X.T, y_pred - y)
   m = len(y)
   weights -= (learning_rate/m) * weights_delta
   return weights
# Define the train function
def linearRegression_train(X, y, learning_rate, num_epochs):
   # Initialize weights and bias
   weights = np.zeros(X.shape[1])
   for epoch in range(num_epochs):
       weights = linearRegression_update_weights(X, y, weights, learning_rate)
       if (epoch % 100 == 0):
           print('epoch: %s, weights: %s' % (epoch, weights))
   return weights
# Define the predict function
def linearRegression_predict(X, weights):
   y_pred = linearRegression_model(X, weights)
   return y_pred
# Define the mean squared error function
def mean_squared_error(y_true, y_pred):
   mse = np.mean(np.power(y_true-y_pred, 2))
   return mse
# Prepare the data
X = data['Height'].values.reshape(-1, 1)
y = data['Weight'].values.reshape(-1, 1)
# Train the model
lr = 0.01
n_epochs = 1000
weights = linearRegression_train(X, y, lr, n_epochs)
# Predict
y_pred = linearRegression_predict(X, weights)
# Evaluate the model
mse = mean_squared_error(y, y_pred)
print('Mean Squared Error: %s' % mse)
# Plot the regression line
plt.scatter(data['Height'], data['Weight'], color='b')
plt.plot(X, y_pred, color='k')
plt.xlabel('Height')
plt.ylabel('Weight')
plt.title('Height vs. Weight')
plt.show()
# Plot the predicted and actual values
plt.scatter(data['Height'], y, color='b', label='Actual')
plt.scatter(data['Height'], y_pred, color='r', label='Predicted')
plt.xlabel('Height')
plt.ylabel('Weight')
plt.title('Actual vs. Predicted')
plt.legend()
plt.show()

i try the same code to run step by step in google colab and i also change the epochs to 62 and run it many times but still the same :

ValueError                                Traceback (most recent call last)
<ipython-input-23-98703406a0a3> in <module>
      2 learning_rate = 0.01
      3 num_epochs = 62
----> 4 weights = linearRegression_train(X, y, learning_rate, num_epochs)

1 frames
<ipython-input-12-8f66dacdd5fc> in linearRegression_update_weights(X, y, weights, learning_rate)
      4    weights_delta = np.dot(X.T, y_pred - y)
      5    m = len(y)
----> 6    weights -= (learning_rate/m) * weights_delta
      7    return weights

ValueError: non-broadcastable output operand with shape (1,) doesn't match the broadcast shape (1,15)

CodePudding user response：

I can reproduce the error message with

In [5]: x=np.array([1])

In [6]: x =np.ones((1,5),int)
---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
Input In [6], in <cell line: 1>()
----> 1 x =np.ones((1,5),int)

ValueError: non-broadcastable output operand with shape (1,) doesn't match the broadcast shape (1,5)

CodePudding user response：

In linearRegression_update_weights, weights.shape == (1,) but weights_delta.shape == (1, 15) so the in-place subtraction fails. The shape of weights_delta is wrong because y_pred.shape == (15,) but y.shape == (15, 1) so (y_pred - y).shape == (15, 15) because of broadcasting. This results in the wrong shape of weights_delta after multiplied by X.T. The fix is to ensure y is a 1-D array to match the shape of y_pred, preventing broadcasting:

y = data['Weight'].values.reshape(-1)