Hello,
After filling out the code (redacted) for exercise 5, and running the test code, I get an error, and notice the accuracy at the 5th decimal point and later doesn’t match. Is this expected behavior?
Your code for compute_cost() is not accurate enough.
These particular tests (by some unreasonable logic) is checking your result to 12 decimal places.
So I suspect your code for compute_cost() is doing some extra mathematical work, which is causing some small roundoff or truncation errors.
1 Like
Thanks for the reply. That is odd; I’ve tried using the various numpy helper functions instead of the regular operators for multiplication/summation. Any other ideas?
I re ran all the notebook cells from the top down, and changing the code to the numpy versions for mult/addition seems to have increased the accuracy, oddly. Thanks
1 Like
The assignment designers assumed you’d be using numpy.
1 Like