Using Python to multiply/divide columns in CSV files

Question

So I have a number of CSV files with 6 columns of numbers in each file. I would like to perform a few operations (multiplication, division etc.) on each column in each of the CSV files using Python.

import csv

r = csv.reader(open('F:\python\sample.csv','rb'))

w = csv.writer(open('F:\python\sample_calib.csv','wb',buffering=0))

for row in r:
            a = (float(row[0])-0.0376)/-0.0717
            b = (float(row[1])-0.0376)/-0.0717
            c = float(row[2])/1000
            d = float(row[3])/1000
            e = float(row[4])/1000000
            f = float(row[5])/0.001178
            w.writerow([a,b,c,d,e,f])

So I am using this small script above to calibrate each row and this works fine for each .csv file. Now all I want to do is to run this script for 200 FILES in one folder. Can some one tell me how should I edit the script and what modules to add?

Answer 1

您需要熟悉csv模块： http ： //docs.python.org/library/csv.html

Answer 2

Assuming you know the arrangement and data type of each column, your best bet would be NumPy and the loadtxt function.

Code to use it would look something like this:

import numpy as np
dtype = np.format_parser(['f4', 'f4', 'i4'], ['col1', 'col2', 'col3'], [])
array = np.loadtxt(path_to_file, dtype, delimiter=',')

Then you can perform operations on an entire column like this.

output = array['col1'] + array['col2']

The f4 and i4 refer to the data type of each column - f4 is a 32 bit floating point number, and i4 is a 32 bit integer. Other options are i8 , f8 , or aN for 64 bit integer, 64 bit float, and N length string.

One caveat - if your data contains strings which include commas, the loadtxt function doesn't handle them well. You'll have to use the csv module as recommended by other posters in that case.