将列表元素转换成数组

Question

I have a list tsv file which I am parsing and want to convert it into an array.

Here is the file format -

jobname1 queue maphours reducehours
jobname2 queue maphours reducehours

code

with open(file.tsv) as tsv:
    line = [elem.strip().split('\t') for elem in tsv]
    vals = np.asarray(line)
    print vals[0]
    print vals[4]

Vals currently returns the following output -

['job1', 'queue', '1.0', '0.0\n']
['job2', 'queue', '1.0', '0.0\n']

I want to convert each element in a row in the entire file to an array object -

vals[0] = job1 vals[1] = queue vals[2] = 1.0 vals[3] = 0.0 

How do i achieve this?

Answer 1

From what I understand you would like to create 2D array in numpy where each row of the file is a row corresponds to the created array, and column in a file is a column in the array. If so, you could do this as follows:

For example, if your data file is:

jobname1    queue   1   3
jobname2    queue   2   4
jobname41   queue   1   1
jobname32   queue   2   2
jobname21   queue   3   4
jobname12   queue   1   6

The following code:

with open(file) as tsv:
    line = [elem.strip().split('\t') for elem in tsv]

vals = np.asarray(line) 

will result in the following vals array:

[['jobname1' 'queue' '1' '3']
 ['jobname2' 'queue' '2' '4']
 ['jobname41' 'queue' '1' '1']
 ['jobname32' 'queue' '2' '2']
 ['jobname21' 'queue' '3' '4']
 ['jobname12' 'queue' '1' '6']]

The get the job names you can do:

print(vals[:,0])
% gives ['jobname1' 'jobname2' 'jobname41' 'jobname32' 'jobname21' 'jobname12']

Or if you want rows containing some job, you can do:

print(vals[np.apply_along_axis(lambda row: row[0] == 'jobname1', 1, vals)])

Answer 2

Are you sure you need an array? @Marcin's answer is more complete if you want a Numpy array.

Python doesn't have an array data structure (there's a list of Python data structures here ). There is a "thin wrapper around the C array" . In order to use the wrapper around the C array, you have to specify a type that the array will hold ( here you'll find a list of typecodes, at the top, and examples at the bottom):

If you want to use a numpy array, this should work:

import numpy as np
myarray = np.asarray(yourList)

adopted from here .