The following code should do what you want:

import numpy as np

f = open('Image Data.txt')

triplets=f.read().split()

for i in range(0,len(triplets)): triplets[i]=triplets[i].split(',')

A=np.array(triplets, dtype=np.uint8)

Thanks for your reply.

Your code is giving the following error in the last line as

"ValueError: setting an array element with a sequence."

Pardon me if I am unable to debug this as I'm quite new to Python. Please reply.

I'm sorry, but I can't reproduce the error. For me, using the given sample file and using Python 2.6.6, it works. I added the script I used. Did you made any changes to the code or did you use a different file?

Oh! My bad. I had some indentation issue. Thanks. Anyways, here's what I'm trying to do. The text file is basically pixel intensities of an RGB image. I managed to reconstruct the image, i.e. form a 30x60x3 matrix. Now, I want to convert it to grayscale. That means that if A(i,j,1)==0 and A(i,j,2) ==0 and A(i,j,0)==0, I want B(i,j) = 0, where B is a 30x60 matrix. I am getting this error :

" File "C:\Users\KU762470\Desktop\converttxt2array.py", line 19

B(i,j) = 0

SyntaxError: can't assign to function call"

FYI, I'm using Python 2.7.6

Appreciate your help.

You have to use square brackets instead of round brackets when assigning values to an array.

You could use "numpy.getfromtxt" method

for an example visit :

http://stackoverflow.com/questions/12319969/how-to-use-numpy-genfromtxt-when-first-column-is-string-and-the-remaining-column

I find that stackoverflow.com is the best place to find answers of this kind, usually the question has been asked and answered already and if it hasn't it will be answered quickly.

Here are some outside the box options: try the Python Image Library (PIL) that will read RGB tuples directly from the image file (e.g., JPG). This library also has image processing for converting RGB to grayscale (or black and white). Also, you may skip Python altogether and use the free ImageMagick software (http://www.imagemagick.org/) that comes with the command line tool 'convert -threshold' that will make a monochrome image from a colored original.

@Tyler Davis, actually, I have the tuples in a .txt file, not the actual image. Anyways, thanks for your reply.

@Hossein: numpy.getfromtxt won't work. The input data is not formatted correctly for this type of read. You could, however, reformat the text file to allow for simpler reads, e.g.:

f = open('Image Data.txt')

d = f.read()

f.close()

import re

d = re.sub(' ', '\n', d) # replace white space character with newline

f = open('Image Data Formatted.txt', 'w')

f.write(d)

f.close()

Now you can easily read the tuples into Python with numpy:

my_data = numpy.loadtxt(

fname='Image Data Formatted.txt',

delimiter = ','

)

shape(my_data) # (1800,3)

import csv

import numpy as np

filename = "Image Data.txt"

reader = csv.reader(open(filename),delimiter=" ")

data = []

for row in reader:

    #each column has one triplet

    for col in row:

        rgb = col.split(",")

        data += [[int(x) for x in rgb]]

npdata = np.array(data,dtype=np.int8)

print(npdata.shape)

some one can help me please.. i need some help in using python code to make an OCR using knn

pandas is great package for dealing with large data sets. It takes some time to learn, but can read data from multiple formats, even Excel spreadsheets.