There are several inconsistencies between the math-formula and py-code

[roveretoa]

There are several inconsistencies between the math-formula and py-code:

(1) Problem 9: Index
In the py-code the index is added by 10, but the right side of the equation shows something else.

(2) Problem 9: Cumsum
The dx lambda parameters should be i, o ?

(3) There are two 'Problem 9'

(4) Problem 19: Matmul
In the py-code the summation is divided by M, but the math-formula is not.

(5) Problem 20: 2D Conv
In the py-code the summation is divided by (kh * kw), but the math-formula is not.

[kshitijsachan]

Also problem 6:
Description reads: $$𝑓(𝑥0)=[𝑥0×0/𝐼,𝑥0×2/𝐼,𝑥0×3/𝐼,…,𝑥0×(𝐼−1)/𝐼]$$

Missing the $$𝑥0×1/𝐼$$ term.

[srush]

Thanks. I'll fix these up.

[ricardoV94]

Shouldn't index j start at 0 in the 1-D convolution formula?

[nilsreck]

There are several inconsistencies between the math-formula and py-code:

(1) Problem 9: Index In the py-code the index is added by 10, but the right side of the equation shows something else.

(2) Problem 9: Cumsum The dx lambda parameters should be i, o ?

(3) There are two 'Problem 9'

(4) Problem 19: Matmul In the py-code the summation is divided by M, but the math-formula is not.

(5) Problem 20: 2D Conv In the py-code the summation is divided by (kh * kw), but the math-formula is not.

In Problem 20, the formula in the py-code is still divided by (kh * kw)