I have a string as follows:
>>> a
'0 911 872.9 354.9 901.9 395.0 904.6 414.0 903.8 400.5'
Now I wish to convert it into an array:
>>> b
array([ 0. , 911. , 872.9, 354.9, 901.9, 395. , 904.6, 414. ,
903.8, 400.5])
What is the most Pythonic way of doing this?
Use numpy.fromstring:
import numpy as np
np.fromstring(a, dtype=float, sep=' ')
Demo:
>>> np.fromstring('0 911 872.9 354.9 901.9 395.0 904.6 414.0 903.8 400.5', dtype=float, sep=' ')
array([ 0. , 911. , 872.9, 354.9, 901.9, 395. , 904.6, 414. , 903.8, 400.5])
The simplest would be to split the string based on white-space followed by mapping the string data to float. You can use imap, if you would not want to create throw away intermediate strings. Also I would suggest to create array's of doubles to reduce precision errors.
Demo
>>> from array import array
>>> from itertools import imap
>>> array('d', imap(float, a.split()))
array('d', [0.0, 911.0, 872.9, 354.9, 901.9, 395.0, 904.6, 414.0, 903.8, 400.5])
>>>
b = array([float(x) for x in a.split()])
b = map(lambda x: float(x), x.split())
numpyand not array