Compute_gradient_test (compute_gradient)

Course Q&A Machine Learning Specialization Supervised ML: Regression and Classification
1

compute_gradient_test

Compute and display cost and gradient with non-zero w

test_w = np.array([0.2, 0.-5])

test_b = -24

dj_db, dj_dw = compute_gradient(X_train, y_train, test_w, test_b)

​

print(‘dj_db at test_w:’, dj_db)

print(‘dj_dw at test_w:’, dj_dw.tolist())

​

​My answer is this

dj_db at test_w: -0.6
dj_dw at test_w: [-44.83135361795273, -44.37384124957207]

But expected outcome is

dj_db at initial w (zeros) -0.5999999999991071
ddj_dw at initial w (zeros): [-44.8313536178737957, -44.37384124953978]

error is AssertionError: Wrong value for dj_db. Expected: 0.28936094 got: 0.4225095475509334.

2

Greetings.

It appears that your code doesn’t work correctly. Please review your code and verify that it implements the equation for dj_db correctly.

A very common problem is using the wrong indentation within any for-loops. Python is very particular about indentation.

3

Thanks for the input. But my value of Compute and display gradient with w initialized to zeroes match with the expected value given.

4

That does not mean your code is correct.
It only means your code works correctly when w and b are zero.

5

Thanks. Any hint that can be helpful

6

Already provided - check your code carefully against the equation you’re implementing.