如何在csv文件中将两个不同长度的列表写入列和行

Question

I have two lists of different length:

list1 = ['a']
list2 = [['apple','banana','grapes']]

What I want in the csv is the following:

col1  col2
a     apple, banana, grapes

So first I tried itertool and izip , but it seems to put only first element of list in the list2 in col2, so it looks like:

col1  col2
a     apple

How can I get the version above, with complete list in list2 as rows?

I used the following code to store the above to csv:

import csv
from itertools import izip

with open('some.csv', 'wb') as f:
    writer = csv.writer(f)
    writer.writerows(izip(list1,sum(list2,[]))

Answer 1

izip will chop the longer list to match the shorter one, so izip(['a'], ['a', 'b', 'c']) actually gives ['a', 'a'] , that is where the problem comes from.

Also generally you would want to use str.join() instead of sum() to convert a list to a string.

From the question I guess you want a csv file delimited by tabs. To solve the problem, first convert list2 to a list of strings:

    >>> list2_str = [','.join(lst) for lst in list2]
    ['apple,banana,grapes']

Then zip list1 and list2_str :

    >>> list3 = zip(list1, list2_str)
    [['a', 'apple,banana,grapes']]

Open a csv writer with 'excel-tab' dialect and write the rows:

    >>> writer = csv.writer(file, dialect='excel-tab')
    >>> writer.writerows(list3)

If instead you want a csv file with ',' as delimiter, simply remove the dialect parameter. Python will correctly quote the second column, producing

    a,"apple,banana,grapes"

in the csv file.

Answer 2

You could use pandas doing something similar to the following:

list1 = ['a']
list2 = [['apple','banana','grapes']]
import pandas as pd
f = pd.DataFrame({'col1':list1, 'col2':list2})
f.to_csv('filename.csv', header=True, index=False)

Answer 3

As @ZdaR pointed out in the comments, commas are the default separator in .csv files, thus, storing your strings separated by a comma might lead to confusion. Instead, you could store your strings separated by a semicolon.

list1 = ['a', 'b']

list2 = [['apple','banana','grapes'], ['foo', 'bar']]

# convert all your lists of strings to single strings separated by a semicolon
list3 = [";".join(li) for li in list2]

list3 looks then as follows:

['apple;banana;grapes', 'foo;bar']

Now you can use your code:

from itertools import izip
import csv

with open('some.csv', 'wb') as f:
    writer = csv.writer(f)
    writer.writerows(izip(list1, list3))

which gives you the following output

a   apple;banana;grapes
b   foo;bar

If you then want to read the file back in again you can do this easily using eg pandas:

import pandas as pd
df = pd.read_csv('some.csv', header=None, names=['col1', 'col2'])

which gives you:

 col1                 col2
0    a  apple;banana;grapes
1    b              foo;bar

Answer 4

The approach depends on if you really want only two columns or more?

If extra columns are required:

import csv
from itertools import izip

list1 = ['a', 'b']
list2 = [['apple','banana','grapes'], ['cherry']]

with open('some.csv', 'wb') as f:
    writer = csv.writer(f)
    writer.writerow(["col1", "col2"])

    for i1, i2 in izip(list1, list2):
        writer.writerow([i1] + i2)

This would give you:

col1,col2
a,apple,banana,grapes
b,cherry

If only two columns are needed, you could switch to using tab characters for your delimiters to allow the commas to separate the list2 entries:

import csv
from itertools import izip

list1 = ['a', 'b']
list2 = [['apple','banana','grapes'], ['cherry']]

with open('some.csv', 'wb') as f:
    writer = csv.writer(f, delimiter='\t')
    writer.writerow(["col1", "col2"])

    for i1, i2 in izip(list1, list2):
        writer.writerow([i1] + [', '.join(i2)])

This would give you:

col1    col2
a   apple, banana, grapes
b   cherry