CODEWITHSUNDEEP
pythonstringnumpymatrix
user7831701
user7831701

Reputation:

Numpy Matrix: Malformed String

I get the following error ValueError: malformed string when using the following code:

import numpy 
l=[]
for x in [0.0,1.0]:
    for y in [0.0,1.0]:
        for z in [0.0,1.0]:
            if x+y+z ==1:
                print x
                pr_matrix = numpy.matrix('x ; y ; z')
                l.append(pr_matrix)

It seems to me that the numpy.matrix function is unable to understand that x,y,z are of type float and not of type string. I say so because the following works fine.

matrix = np.matrix('1 2 2;2 3 1')

Upvotes: 2

Views: 450

Answers (1)

percusse
percusse

Reputation: 3106

Do not use numpy.matrix. It's deprecated and will be removed in the future NumPy releases. Instead you can use a simple comprehension

import itertools, numpy as np
arr=np.array([[[x],[y],[z]] for x,y,z in itertools.product(*[[1,0]]*3) if x+y+z==1])[:,:,0]

Here, itertools.product is creating the cartesian product [0, 1]x[0, 1]x[0, 1] which is what you do with 3 nested loops. To give this argument to the itertools.product we basically first create three copies of [0,1] in an iterable.

Then with a star argument (*) before this iterable, we instruct to interpret the elements of this iterable as separate arguments so we get three distinct arguments. This produces the cartesian product 

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

let's call this A. Then we do a list comprehension over A. For each element in A which is a 3-tuple x,y,z we form a column and add it to the list created by the list comprehension if their sum is 1.

The result is then

[[[1], [0], [0]], [[0], [1], [0]], [[0], [0], [1]]]

and then an NumPy array is created but notice it is 3D. so we take the relevant part with a slice. You can also use np.squeeze for this.

Upvotes: 2

Related Questions