What's the fastest way to identify elements in one numpy array, and use the index to look up elements in a second array?

Question

I have two arrays. The first being an array of IDs, and the second being values for for those ids. I am trying to looks up the values for each unique ID from array 1 based on the index of the matching arrays. The values are classified integers 1-16. My current attempt looks like:

for id in np.unique(ID_array):
    temp_array = np.where(ID_array == id, 1, 0)
    test_array = Value_array * temp_array
    count_array = np.where(test_array != 0)

Test_data:

ID_array = np.random.randint(1,10,(5,5))
Value_array = np.random.randint(1,16,(5,5))

From there I am just determining the percent distribution of each value class for the unique IDS.

What I am wondering, is there a faster way to identify the matching values in the second array other than creating a Boolean mask?