Write numpy arrays to files using csv

Question

I have a list of .asc files with geographic coordinates that I would like to load into a .csv file for further analysis in Excel. Each .asc file contains approximately 2000 rows.

list_files = ['./p2c0765f029.asc',..., './p2c0781f029.asc']


with open('output_file.csv', 'a+') as outfile:
    for file in list_files:
        file_in = loadtxt(file, usecols = (1,2), comments = "#")      
        csvwriter = csv.writer(outfile, delimiter = ' ')
        csvwriter.writerow('%s\n' % file_in)

However this outputs a file in which each array is stored in one row (when I open the file in Excel).

[ [ " " - 8 . 7 9 8 4 5 4 5 0 e + 0 1 " " " " " " 9 . 5 5 1 5 5 4 8 0 e + 0 1 " " " " ]]

How could I implement this so each line in the numpy arrays is written in one row? Also I would like to get rid of " and [] and output only the coordinates.

You have numpy and you want to do further anaysis in excel? Gee... Let me translate how it sounds. I have an automatic machine gun, I'd like to hunting with a spear. — oz123
– oz123, Commented Jul 22, 2014 at 9:08
Does docs.scipy.org/doc/numpy/reference/generated/numpy.savetxt.html not do what you want? — EdChum
– EdChum, Commented Jul 22, 2014 at 9:10
@Oz123 think of it more as a Bayonet :). But seriously: I guess if you are trained in using pivot tables, there is a lot of analysis you can do faster and easier in excel. — Chronial
– Chronial, Commented Jul 22, 2014 at 9:30

Pau · Accepted Answer · 2014-07-22 09:41:29Z

3

Using numpy.savetxt I get the output I needed:

with open('output_file.txt', 'a+') as outfile:
     for file in list_files:
         file_in = loadtxt(file, usecols = (1,2), comments = "#")      
         np.savetxt(outfile, file_in, delimiter=' ', newline='\n')

Not sure if it's the most pythonic way, but it works.

answered Jul 22, 2014 at 9:41

Pau

932 silver badges12 bronze badges

Sign up to request clarification or add additional context in comments.

2 Comments

EdChum Over a year ago

To me pythonic is not an issue here, numpy has provided a method for loading text data and provided a method for saving it, therefore this would be the preferred method. Incidentally Pandas is much faster at loading csv data and probably writing it out, it has a friendlier api than numpy

Chronial Over a year ago

Yes, using the api is always pythonic :).

Collectives™ on Stack Overflow

Write numpy arrays to files using csv

1 Answer 1

2 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

2 Comments

Your Answer

Sign up or log in

Post as a guest

Related