# DLS W3 assignment question

Hello,
In the exercise 2 it is asking for the size of the input layer n_x. From the previous exercise I can see the the number of training samples x = 400, therefore m = 400. From the lecture I understood that n_x = number of training samples m, therefore it should be 400. However the expected outcome is:
Expected output
The size of the input layer is: n_x = 5
The size of the hidden layer is: n_h = 4
The size of the output layer is: n_y = 2

``I don't understand why n_x is 5, if someone could explain would be very helpful. Thank you!``

The point is that we are writing code which is â€śgeneralâ€ť, meaning that it can handle data with different numbers of features and networks with different number of neurons. That is just a test case that is checking that the code works. Note that `n_x` is the number of features, not the number of samples. The code should work with `n_x = 5`, just as well as it works with `n_x = 2`, right?

BTW you filed this question under a generic category. I will move it to the DLS Course 1 category by using the little â€śedit pencilâ€ť on the title.

2 Likes

Hello,

The size of the input layer of your model (or any model) doesnâ€™t depend on the number of samples. It depends on the number of features.

In this case we got two features (2 coordinates), so `n_x` should be 2. The code is not hard-coding the value 2, but `X.shape[0]`, which is effectively the same thing.

Now the print is a bit misleading because it actually prints the number of features from `t_X` from the test case.

Just remember n_x = number of features, thatâ€™s all!

2 Likes

Thank you Paul and Mathieu for the clarification. I used shape.X[0] command for n_x which resulted equal the number of parameters - 2, however it is still not accepted:
AssertionError: Wrong result. Expected (5, 4, 2) got (2, 4, 2).

To make sure I create the question under the correct category next time - I usually go to the AI Discussions tab on the left, then - create new topic, then I just type the name of the course and the week. I didnâ€™t see where I can specify the course or maybe I am starting the topic in the wrong section to begin with?
Thank you

1 Like

It sounds like your mistake is referencing the global variable `shape_X` from within the function. You should only reference the parameters passed to your function. In these courses, it is always a mistake to reference global variables from the body of a function.

There are categories for the DLS courses, so thatâ€™s the way to open a thread about DLS.

1 Like

Start from the Course Q &A section. Not AI Discussions.

2 Likes