C1_W3_Assignment forward_propagation bug

Course Q&A Math for Machine Learning Linear Algebra for Machine Learning and Data Sc...
1

Found a bug in. I was getting this error when running the Multiple Linear Regression.

---------------------------------------------------------------------------
AssertionError                            Traceback (most recent call last)
<ipython-input-43-a52416c0c3bf> in <module>
      1 ### START CODE HERE ### (~ 1 line of code)
----> 2 parameters_multi = nn_model(X_multi_norm, Y_multi_norm, num_iterations=100, print_cost=False)
      3 ### END CODE HERE ###
      4 
      5 print("W = " + str(parameters_multi["W"]))

<ipython-input-20-105b6356d917> in nn_model(X, Y, num_iterations, print_cost)
     26         ### START CODE HERE ### (~ 2 lines of code)
     27         # Forward propagation. Inputs: "X, parameters". Outputs: "Y_hat".
---> 28         Y_hat = forward_propagation(X, parameters)
     29 
     30         # Cost function. Inputs: "Y_hat, Y". Outputs: "cost".

<ipython-input-42-6ea40968b18a> in forward_propagation(X, parameters)
     22     ### END CODE HERE ###
     23 
---> 24     assert(Y_hat.shape == (n_x, X.shape[1]))
     25 
     26     return Y_hat

AssertionError:

n_X is not computed in the Multiple Linear Regression section and is not given to forward_propagation as a parameter. This variable holds the value (1) from the Single Single Linear Regression code.

I commented out

assert(Y_hat.shape == (n_x, X.shape[1]))

in forward_propagation to get it to work.

2

Hi!

Even though the parameter is not passed, you have to think the neural net as a sequence of functions and n_x it is used to generate the W.