Numpy array of strings into an array of integers-CodePudding

I have the following array:

pattern = array([['[0, 0, 1, 0, 0]'],
       ['[0, 1, 1, 1, 1]'],
       ['[0, 1, 1, 1, 0]'],
       ['[0, 0, 1, 1, 1]'],
       ['[0, 0, 0, 1, 1]'],
       ['[0, 0, 1, 0, 1]'],
       ['[0, 0, 0, 0, 1]'],
       ['[1, 0, 1, 0, 0]'],
       ['[0, 1, 0, 1, 1]'],
       ['[0, 0, 1, 1, 0]'],
       ['[1, 1, 1, 1, 1]'],
       ['[1, 1, 1, 1, 0]']], dtype='<U15')

and I want to get it in non-string format as the following:

import numpy
my_array = numpy.array([[0, 0, 1, 0, 0],
                        [0, 1, 1, 1, 1],
                        [0, 1, 1, 1, 0],
                        [0, 0, 1, 1, 1],
                        [0, 0, 0, 1, 1],
                        [0, 0, 1, 0, 1],
                        [0, 0, 0, 0, 1],
                        [1, 0, 1, 0, 0],
                        [0, 1, 0, 1, 1],
                        [0, 0, 1, 1, 0],
                        [1, 1, 1, 1, 1],
                        [1, 1, 1, 1, 0]
                        ])

Any idea on how to do it non-manually?

CodePudding user response：

You can use the ast.literal_eval() function from the ast module to evaluate the strings in the array as Python literals and convert them to lists. Then you can use the numpy.array() function to create a NumPy array from the lists.

Here's an example of how you can do this:

import ast
import numpy

pattern = [['[0, 0, 1, 0, 0]'],
           ['[0, 1, 1, 1, 1]'],
           ['[0, 1, 1, 1, 0]'],
           ['[0, 0, 1, 1, 1]'],
           ['[0, 0, 0, 1, 1]'],
           ['[0, 0, 1, 0, 1]'],
           ['[0, 0, 0, 0, 1]'],
           ['[1, 0, 1, 0, 0]'],
           ['[0, 1, 0, 1, 1]'],
           ['[0, 0, 1, 1, 0]'],
           ['[1, 1, 1, 1, 1]'],
           ['[1, 1, 1, 1, 0]']
           ]

# Evaluate the strings as Python literals and convert them to lists
pattern = [ast.literal_eval(x[0]) for x in pattern]

# Create a NumPy array from the lists
pattern = numpy.array(pattern)

print(pattern)

This will output the following NumPy array:

array([[0, 0, 1, 0, 0],
       [0, 1, 1, 1, 1],
       [0, 1, 1, 1, 0],
       [0, 0, 1, 1, 1],
       [0, 0, 0, 1, 1],
       [0, 0, 1, 0, 1],
       [0, 0, 0, 0, 1],
       [1, 0, 1, 0, 0],
       [0, 1, 0, 1, 1],
       [0, 0, 1, 1, 0],
       [1, 1, 1, 1, 1],
       [1, 1, 1, 1, 0]])

CodePudding user response：

Using numpy string operations to strip brackets ([]), splitting on comma and recast into an array with int dtype is possible:

np.array(np.char.split(np.char.strip(pattern[:, 0], '[]'), ', ').tolist(), 'int')

but a list comprehension where you do the same things using python string methods is much easier to read (and faster as well) imo.

np.array([row[0][1:-1].split(', ') for row in pattern], dtype='int')


# array([[0, 0, 1, 0, 0],
#        [0, 1, 1, 1, 1],
#        [0, 1, 1, 1, 0],
#        [0, 0, 1, 1, 1],
#        [0, 0, 0, 1, 1],
#        [0, 0, 1, 0, 1],
#        [0, 0, 0, 0, 1],
#        [1, 0, 1, 0, 0],
#        [0, 1, 0, 1, 1],
#        [0, 0, 1, 1, 0],
#        [1, 1, 1, 1, 1],
#        [1, 1, 1, 1, 0]])