Week 1, Assignment 1, Variable Naming

Hello Instructors,

I am working on the first assignment of week 1 for this course and I find the variable naming is very confusing to me.

For the W parameters, its dimension is defined as:
(f, f, n_C_prev, n_C) = W.shape

The first 2 f are easy to understand, they are the size of the kernel.
I am often confused about the latter dimensions.

I think n_C_prev is the number of channels, which must be equal to the number of channels of the input image. And n_C is the number of filters. Is that right?

If yes, it is difficult to read the code as both variables share the same “n_C” prefix. Perhaps the n_C variable should be renamed as n_F.

I hope you clear my doubt. Thank you in advance.

Best Regards,
Tri Tam

Hey @geektom n_C = number of channels of the current layer
n_C_prev = number of channels of the previous layer.