Combine cells from one column into one string

Question

df = pd.read_csv(filename.csv)
corpus = df.corpus

How can I combine series of text strings (from one column) into a list?

from column 'corpus': row 1: Hail Mary.
row 2: Hi Bob.
row 3: Hey Sue.

into [Hail Mary. Hi Bob. Hey Sue.]

Looking for a list with len(list)=1.

A column of text in what? You haven't shown what the file type is, let alone how you're reading it. I also don't understand "Looking for a list with `len(list)=1`." — roganjosh, Aug 23 '18 at 16:16
Please show the code with the construct which stores your text. — Yunnosch, Aug 23 '18 at 16:18
@mad Tried that already, returns a str object. I'd like a list. — spacedustpi, Aug 23 '18 at 16:19
how will it return a string if encapsulated by `[` and `]` @spacedustpi — mad_, Aug 23 '18 at 16:20
And that keeps you from showing any code to help people understand and answer your question? — Yunnosch, Aug 23 '18 at 16:23
Possible duplicate of [Combine two columns of text in dataframe in pandas/python](https://stackoverflow.com/questions/19377969/combine-two-columns-of-text-in-dataframe-in-pandas-python) — mad_, Aug 23 '18 at 16:24
@mad, those are columns, not rows they are working with in that link. — spacedustpi, Aug 23 '18 at 16:36
@spacedustpi Thanks for growing the chat area without providing the https://stackoverflow.com/help/mcve — mad_, Aug 23 '18 at 16:38
Okay, I will post all the techniques that I tried. I was just trying to keep all the excess off as much as possible, but it seems this is not as easy as I thought. Which I guess boosts my ego. — spacedustpi, Aug 23 '18 at 16:40

Joe · Accepted Answer · 2018-08-23T17:19:05.947

2

If I understood you correctly:

df = pd.read_csv(your_file)
l = [' '.join(df['col'])]

Input:

          col
0  Hail Mary.
1     Hi Bob.
2    Hey Sue.

Output:

['Hail Mary. Hi Bob. Hey Sue.']

edited Aug 23 '18 at 17:19

answered Aug 23 '18 at 16:55

Joe

12,057
5
39
55

It actually works too with the square brackets and omitting '.tolist()' – spacedustpi Aug 23 '18 at 17:17

chriscberks · Answer 2 · 2018-08-23T17:20:50.930

1

import pandas as pd

df = pd.read_csv('example.csv')

result = [' '.join([row for row in df['column_name']])]

Output of result:

['Hail Mary. Hi Bob. Hey Sue.']

edited Aug 23 '18 at 17:20

answered Aug 23 '18 at 16:49

chriscberks

425
2
8

This technique yields a string object. Not a list. – spacedustpi Aug 23 '18 at 17:07
It returns a `list` containing a single `string`. The brackets `[...]` around the `join` give you a `list`. Hope this helps! – chriscberks Aug 23 '18 at 17:15
1

Yes it does. Even simpler is result = [' '.join(df['column_name'])] -thanks. – spacedustpi Aug 23 '18 at 17:23

score 0 · Answer 3 · answered Aug 23 '18 at 16:41

0

test.csv:

Hail Mary. Hi Bob. Hey Sue.

python:

import csv
data = []
with open('test.csv','rb') as csvfile:
    for row in csvfile:
        data.append(row.strip())
print data

output: ['Hail Mary.', 'Hi Bob.', 'Hey Sue.']

answered Aug 23 '18 at 16:41

sur.la.route

440
3
11

1

As per OP's comments, the combine should work on rows and not columns. `row` in your case will contain all the columns in the single row. Hopefully, you can edit your answer once he posts something otherwise it will be downvoted – mad_ Aug 23 '18 at 16:45
@Christopher, this sort of works, but I'd skip the first row because it brings in the column labels into the list. In addition, the list length comes out to 5, I was looking for length of 1. – spacedustpi Aug 23 '18 at 17:35

score -1 · Answer 4 · answered Aug 23 '18 at 16:27

-1

    my_list = []

    for i in range(1,ws.max_row):
        if(len(my_list) == 0):
            my_list.append(ws.cell(row=i,column=1).value)   #assuming the value in 1st column
        else:
            my_list[0] += ws.cell(row=i,column=1).value

    for i in my_list:
        print(i)

answered Aug 23 '18 at 16:27

Somesh

82
8

1

This isn't pandas syntax, it looks like you're using an Excel reader. I don't understand how this deals with the OP's problem – roganjosh Aug 23 '18 at 16:28
I'm working with millions of lines, so I would not use this option even if we worked the kinks out. For loops are notoriously slow with big data sets. – spacedustpi Aug 23 '18 at 17:37

Combine cells from one column into one string

4 Answers4