CODEWITHSUNDEEP
pythonnumpy
Muhammad Imran
Muhammad Imran

Reputation: 9

Changing data type in a Matrix

This is my code

L = np.random.choice([-1,1], size = [2,2])

np.putmask(L, L == -1, "+")

print(L)

I get this error invalid literal for int() with base 10: '+' . Any solution for this problem.

Upvotes: 0

Views: 73

Answers (2)

hpaulj
hpaulj

Reputation: 231395

In [227]: np.random.choice([-1,1], size=[2,2])                                  
Out[227]: 
array([[ 1,  1],
       [-1,  1]])

Change the numbers to strings:

In [229]: x=np.random.choice(["-1","1"], size=[2,2])                            
In [230]: x                                                                     
Out[230]: 
array([['-1', '1'],
       ['-1', '-1']], dtype='<U2')

Note the dtype, 'U2'. Now we can replace some of those strings with another string:

In [231]: np.putmask(x,x=="-1", '+')                                            
In [232]: x                                                                     
Out[232]: 
array([['+', '1'],
       ['+', '+']], dtype='<U2')

I could do string operations on elements of such an array:

In [233]: [''.join(row) for row in x.tolist()]                                  
Out[233]: ['+1', '++']

Upvotes: 0

epattaro
epattaro

Reputation: 2428

It looks like you are trying to convert all negative terms to positive.

The issue here is with your use of the 'np.putmask' function

From the documentation example:

>>> x = np.arange(6).reshape(2, 3)
>>> np.putmask(x, x>2, x**2)
>>> x
array([[ 0,  1,  2],
       [ 9, 16, 25]])

Hence, for your code to work you should use:

>>>L = np.random.choice([-1,1], size = [2,2])

>>>np.putmask(L, L == -1, np.abs(L))

>>>print(L)

[[1 1]
 [1 1]]

Now, if instead, what you actually want is to replace it by the symbol "+" You have to change the dtype of the array.

Since np.random.choice works with a fixed input, it uses the same data type when building the random array! Hence, , you can change it after its built, or change the input, like this:

L = np.random.choice([-1,1], size = [2,2])
L = np.array(L, dtype='object')
np.putmask(L, L == -1, "+")

or like that:

L = np.random.choice(np.array([-1,1],dtype='object'), size = [2,2])
np.putmask(L, L == -1, "+")

print(L)

hope it helps!

Upvotes: 1

Related Questions