Unit test in exercise 2 "initialize_parameters_test" not matching random.seed(2)

Course Q&A Deep Learning Specialization Neural Networks and Deep Learning
1

I dont think the numbers in the unit test “initialize_parameters_test” match the numbers generated by np.random.seed(2) . Are you sure this unit test is correct?

def initialize_parameters_test(target):
np.random.seed(2)
n_x, n_h, n_y = 3, 5, 2

expected_output = {'W1': np.array(
    [[-0.00416758, -0.00056267, -0.02136196],
     [ 0.01640271, -0.01793436, -0.00841747],
     [ 0.00502881, -0.01245288, -0.01057952],
     [-0.00909008,  0.00551454,  0.02292208],
     [ 0.00041539, -0.01117925,  0.00539058]]), 
                   'b1': np.array([[0.], [0.], [0.], [0.], [0.]]), 
                   'W2': np.array([[-5.96159700e-03, -1.91304965e-04,  1.17500122e-02,
    -7.47870949e-03,  9.02525097e-05],
   [-8.78107893e-03, -1.56434170e-03,  2.56570452e-03,
    -9.88779049e-03, -3.38821966e-03]]), 
                   'b2': np.array([[0.], [0.]])}
2

Setting the seed value to 2 ensures that every time you run the code, you will get the same set of random numbers.

3

do you get the same random numbers which are hard coded in the unit test example though when using seed 2?

4

Yes, that is the idea.

5

sorry, i just cant get this unit test to pass.

np.random.seed(2)
W1=np.random.rand(5,3)*0.01

gives:
array([[0.00435995, 0.00025926, 0.00549662],
[0.00435322, 0.00420368, 0.00330335],
[0.00204649, 0.00619271, 0.00299655],
[0.00266827, 0.00621134, 0.00529142],
[0.0013458 , 0.00513578, 0.0018444 ]])

which is not the same as the hard coded values in the unit test. what have i done wrong?

6

Share your full error.

7

ah this is my error - in the example i gave above i used the incorrect random function.

i used np.random.rand and not np.random.randn

8

Glad to know that you solved it on your own.