Python loop through text and set numpy array index

Question

Given a block of text with matrix rows and columns separated by commas and semicolons, I want to parse the text and set the indices of numpy arrays. Here is the code with the variable 'matrixText' representing the base text.

I first create the matrices and then split the text by semicolons and then by commas. I loop through the split text and set each index. However with the text ...

1,2,3;4,5,6;7,8,9

I get the result

7,7,7;8,8,8;9,9,9

temp1=matrixText.split(';')
temp2=temp1[0].split(',')
rows=len(temp1)
columns=len(temp2)
rA=np.zeros((rows, columns))
arrayText=matrixText.split(';')
rowText=range(len(arrayText))
for rowIndex, rowItem in enumerate(arrayText):
    rowText[rowIndex]=arrayText[rowIndex].split(',')
    for colIndex, colItem in enumerate(rowText[rowIndex]):
        rA[[rowIndex, colIndex]]=rowText[rowIndex][colIndex]

I thought that by setting each index, I would avoid any copy by reference issues.

To provide more info, in the first iteration, the 0,0 index is set to 1 and the output of that is then 1,1,1;0,0,0;0,0,0 which I can't figure out since setting one index in the numpy array sets three.

In the second iteration, the index 0-1 is set to 2 and the result is then 2,2,2;2,2,2;0,0,0

The third iteration sets 0-2 to 3 but the result is 3,3,3;2,2,2;3,3,3

Any suggestions?

Answer 1

You can (ab-) use the matrix constructor plus the A property

np.matrix('1,2,3;4,5,6;7,8,9').A

Output:

array([[1, 2, 3],
       [4, 5, 6],
       [7, 8, 9]])

Answer 2

matrixText = '1,2,3;4,5,6;7,8,9'
temp1=matrixText.split(';')
temp2=temp1[0].split(',')
rows=len(temp1)
columns=len(temp2)
rA=np.empty((rows, columns),dtype=np.int)
for n, line in enumerate(temp1):         
    rA[n,:]=line.split(',')

Answer 3

Using a nested list-comprehension :

Having defined:

s = "1,2,3;4,5,6;7,8,9"

we can use a nice one-liner :

np.array([[int(c) for c in r.split(",")] for r in s.split(";")])

which would give the following array :

array([[1, 2, 3],
       [4, 5, 6],
       [7, 8, 9]])

Answer 4

Another one-liner (+1 import):

from io import StringIO
rA = np.loadtxt(StringIO(matrixText.replace(';','\n')), delimiter=',')

Answer 5

So, the problem here was that the dual brackets in rA[[rowIndex, colIndex]] caused every cell in the row to be set. This should be rA[rowIndex, colIndex]