Python: Split CSV file according to first character of the first column

问题

I have a series of large CSV files "basename.csv" like:

B1,3,5,6

B2,2,1,5

B3,1,9,0

C1,4,7,9

C2,1,9,3

C3,8,5,2

I would like to split them into different files like:

basename_B.csv

B1,3,5,6

B2,2,1,5

B3,1,9,0

basename_C.csv

C1,4,7,9

C2,1,9,3

C3,8,5,2

I have already done similar things in the past with for loops and ifs, but I was wondering if there is a more efficient way of doing this with Pandas or whatever.

SOLUTION

Adapting the solution from @chthonicdaemon and @jezrael, I came up with this:

def split_csv():
    for dfile in glob.glob('*.csv'):
        df = pd.read_csv(dfile, header=None)
        for letter, group in df.groupby(df[0].str[0]):
            group.to_csv((os.path.splitext(dfile)[0]) + '_{}.csv'.format(letter), index=False, header=False)

split_csv()

回答1:

Here's a simple application of groupby:

df = pandas.read_csv('basename.csv', header=None)

def firstletter(index):
    firstentry = df.ix[index, 0]
    return firstentry[0]

for letter, group in df.groupby(firstletter):
    group.to_csv('basename_{}.csv'.format(letter))

Or, incorporating @jezrael's use of grouping by the explicit contents of the columns:

for letter, group in df.groupby(df[0].str[0]):
    group.to_csv('basename_{}.csv'.format(letter))

回答2:

I think you can create MultiIndex and then groupby by first level of index and use to_csv:

import pandas as pd
from pandas.compat import StringIO

temp=u"""B1,3,5,6
B2,2,1,5
B3,1,9,0
C1,4,7,9
C2,1,9,3
C3,8,5,2"""
#after testing replace StringIO(temp) to filename
df = pd.read_csv(StringIO(temp), header=None)
print (df)
    0  1  2  3
0  B1  3  5  6
1  B2  2  1  5
2  B3  1  9  0
3  C1  4  7  9
4  C2  1  9  3
5  C3  8  5  2

Another similar solution as another answer:

for letter, g in df.groupby([df.iloc[:, 0].str[0]]):
    #print (letter)
    #print (g)
    g.to_csv('basename_{}.csv'.format(letter))

回答3:

I tested this code and should meet the requirements of your request. example.csv would be the csv inputfile.

with open ('example.csv') as f:
    r = f.readlines()

for i in range(len(r)):
    row = r[i]
    letter = r[i].split(',')[0][0]
    filename = "basename_"+letter.upper()+".csv"
    with open(filename,'a') as f:
        f.write(row)

来源：https://stackoverflow.com/questions/40789383/python-split-csv-file-according-to-first-character-of-the-first-column

标签

python

csv

pandas

split

multiple-columns