使用numpy.genfromtxt读取单行CSV

Question

I am using the following script to read a file from standard input using numpy.

#!/usr/bin/env python
import numpy as np
import sys

data = np.genfromtxt(sys.stdin, delimiter=",")
print data.shape
print data

This works correctly for files that have more than 1 line. But fails to work for this file:

1,2,2,2,2,2,1,1,1

I am running it like this

$ cat input-file.txt | ./test.py

The output is as follows:

(9,)
[ 1.  2.  2.  2.  2.  2.  1.  1.  1.]

It should have shape (,9). Does anyone know how to fix it?

Answer 1

Force it into a 2-dimensional array:

data = np.genfromtxt(sys.stdin, delimiter=",")
if len(data.shape) == 1:
    data = np.array([data])

Answer 2

You could use csv which works in all cases:

import csv

reader=csv.reader(sys.stdin)
for row in reader:
    print row

Testing from Bash:

$ printf '1,2,3,4' | ./test.py
['1', '2', '3', '4']
$ printf '1,2,3,4\n5,6,7,8' | ./test.py
['1', '2', '3', '4']
['5', '6', '7', '8']