python - initialize objects in numpy matrix -

i have numpy matrix filled unique objects. creating list of list , converting numpy array (see code below, under workaround). using because want use slicing access elements in matrix

i wondering if there better way create such matrix.

import random import numpy np  class randomcell(object):     def __init__(self):         self.value = random.randint(0, 10)     def __repr__(self):         return str(self.value)  # workaround temp_matrix = [[randomcell() row in range(3)] col in range(3)] workaround_matrix = np.array(temp_matrix) 

edit: want create matrix of objects not generate matrix of random numbers

your method of building array list of lists fine. option be

arr = np.array([randomcell() item in range(9)]).reshape(3,3) 

usually, save memory use np.fromiter build array iterator. however, since array has dtype object, unfortunately np.fromiter not option in case.