-1

I have a 2d array and i want to write it to a file, the array looks almost like this:

>>print (arr)
[0 0 20 
0 5 520 
2 0 720
.... 
8 -20 150
0 10 10]

when i tried to write it to a file it was saved as the last output, here what i used:

ff = open('output.txt', 'w')
ff.write(arr)

this was the result in the file

 [0 0 20 
 0 5 520 
 2 0 720
 .... 
 8 -20 150
 0 10 10]

i saw a solution in another question in this website but still i have a problem,

np.ndarray.tofile(arr,"output.txt",'\n','%s')

the output in the file was like this :

0 0 20 0 50 ...

and so on till the end of the array

i want the output to look like this:

 0 0 20 
 0 5 520 
 2 0 720
 .
 .
 .
 8 -20 150
 0 10 10
ali_m
  • 71,714
  • 23
  • 223
  • 298
Prince
  • 123
  • 1
  • 2
  • 9
  • 1
    You want [`np.savetxt`](http://docs.scipy.org/doc/numpy/reference/generated/numpy.savetxt.html) – ali_m Nov 03 '15 at 20:42
  • @ali_m i tried np.savetxt but it takes about 26 second whcih is way too large time – Prince Nov 03 '15 at 21:03
  • i wanted to try this one `import numpy as np x = np.arange(20).reshape((4,5)) np.savetxt('test.txt', x)` but the problem that i didn't know how to arrange and reshape my array – Prince Nov 03 '15 at 21:04
  • 2
    Saving to text is only useful if you need your array to be in a human-readable format. If your array takes 26 seconds to save then I can't imagine that many humans would ever want to read it... If you want performance then you should save as a `.npy` file (e.g. using [`np.save`](http://docs.scipy.org/doc/numpy/reference/generated/numpy.save.html)) or as a raw binary (`array.tofile()`). – ali_m Nov 03 '15 at 21:06
  • It sounds as though your real problem is that you don't understand how to reshape numpy arrays. The output in your question looks like a 1D array (or maybe even a list) rather than a 2D array as your question title states. At any rate, saving numpy arrays to text files is an issue that has been covered many times before on SO, so I'm voting to close. – ali_m Nov 03 '15 at 21:47

1 Answers1

0

You could split your array into chunks and write each chunk separately.

Something like this should get you started:

myarr = [1,10,100,2,20,200,3,30,300,4,40]

def split_to_chunks(myarray, e):
    return (myarray[i:i+e] for i in xrange(0, len(myarray), e))

for i in split_to_chunks(myarr, 3):
    # etc
Andrew Calder
  • 451
  • 4
  • 8
  • That's a list, not a numpy array (although for all I can tell the OP might be using a list as well) – ali_m Nov 03 '15 at 21:41