Comparing two numpy arrays

Sahil Goyal 0
Viewed 39

Initializing the empty numpy array

y=np.empty((2,2),dtype=np.matrix)
b=np.empty((2,2),dtype=np.matrix)

Assigning values to above arrays

b[0][0]=np.mat([
[67,57],
[19,56]])

b[0][1]=np.mat([
[7,58],
[9,46]])  

b[1][0]=np.mat([
[77,47],
[34,34]])

b[1][1]=np.mat([
[2,66],
[78,45]])

y[0][0]=np.mat([
[67,57],
[19,56]])

y[0][1]=np.mat([
[7,58],
[9,46]]) 

 y[1][0]=np.mat([
[77,47],
[34,34]])

y[1][1]=np.mat([
[2,66],
[78,45]])

Printing the array

print(y)
print(b)

The y and b arrays are equal and it should print True but instead it is printing False

print(np.array_equal(y,b))
print(y==b)
numpy-slicing matrix numpy arrays python
1 Answers

Firstly, as Michael Szczesny mentioned, np.mat is deprecated and you should avoid it. Object arrays are also kinda iffy but can serve a purpose sometimes. Although to answer the actual question:

To check for equality between these arrays of arrays you need to check that the inner arrays are equal and then go from there. To do this you can trick numpy into applying np.array_equal element wise and then check with np.all

import numpy as np
deep_equal = np.vectorize(np.array_equal)
def inner_array_equal(a,b): return np.all(deep_equal(a,b))


a = np.empty(2,dtype=object)
b = np.empty(2,dtype=object)

a[0] = np.array([1,2])
a[1] = np.array([3,4])

b[0] = np.array([1,2])
b[1] = np.array([3,4])

print(a,b)

print(inner_array_equal(a,b))

Notice that I switched your code to be explicit object arrays (although dtype=np.matrix will do the same thing) and filled them with np.array instead.

Simon Tartakovksy0
Related