I am trying to insert a string into a integer based matrix in numpy matrix. I am wondering if their is a way to change the datatype of the index that I want to store the string in?
I have tried to use the .astype() function but had no luck in doing so. This is what I have tried to do
c = np.array([0])
c.resize((3,3))
c.fill(0)
c.astype(str)
c.itemset(2, 0, 'S')
This is what I am trying to have my output look like:
OutPut:
[[0 0 S]
[0 0 0]
[0 0 0]]
You need to set the dtype to object in order to allow multiple data types inside an array:
c = np.array([0])
c.resize((3,3))
c.fill(0)
c = c.astype('O')
c.itemset(2, 0, 'S')
c
>> array([[0, 0, 0],
[0, 0, 0],
['S', 0, 0]], dtype=object)
Sidenote: numpy arrays are not meant to be multiple type and this will likely be underperformant
Make your c (a simpler way):
In [377]: c = np.zeros((3,3), int)
In [378]: c
Out[378]:
array([[0, 0, 0],
[0, 0, 0],
[0, 0, 0]])
Trying assign a string to an integer dtype results in an error - it can't convert the string to an integer:
In [379]: c[0,2] = 'S'
---------------------------------------------------------------------------
ValueError Traceback (most recent call last)
<ipython-input-379-d41cf66ff6a1> in <module>
----> 1 c[0,2] = 'S'
ValueError: invalid literal for int() with base 10: 'S'
astype does not work in-place:
In [380]: c.astype(str)
Out[380]:
array([['0', '0', '0'],
['0', '0', '0'],
['0', '0', '0']], dtype='<U21')
In [381]: c
Out[381]:
array([[0, 0, 0],
[0, 0, 0],
[0, 0, 0]])
Make a new array with the string dtype:
In [382]: c1 = c.astype(str)
Now we can assign the new string value:
In [383]: c1[0,2] = 'S'
In [384]: c1
Out[384]:
array([['0', '0', 'S'],
['0', '0', '0'],
['0', '0', '0']], dtype='<U21')
