Reading csv file as dictionary using pandas

Question

I have the foll. csv with 1st row as header:

A      B
test    23
try     34

I want to read in this as a dictionary, so doing this:

dt = pandas.read_csv('file.csv').to_dict()

However, this reads in the header row as key. I want the values in column 'A' to be the keys. How do I do that i.e. get answer like this:

{'test':'23', 'try':'34'}

Thanks for clarifying this, I want {'test':'23', 'try':'34'}. Updated question — user308827, Nov 10 '15 at 00:45
And your delimiter is `\t`? Generally, `csv` files are `1, 2, 3\n4, 5, 6`. Commas and newlines to separate columns and rows. — Zizouz212, Nov 10 '15 at 00:46
Possible duplicate of [Convert Pandas DataFrame to dictionary](http://stackoverflow.com/questions/26716616/convert-pandas-dataframe-to-dictionary) — R Nar, Nov 10 '15 at 00:47
@RNar Nah, they look like completely different questions with different objects. — Zizouz212, Nov 10 '15 at 00:49
@Zizouz212 i will have to disagree with you. the inherent problem is that he wants to have a dict where the keys are the values in a column rather than the columns themselves. the answer to that is to set that column is the index of the dataframe then call `to_dict('list')` — R Nar, Nov 10 '15 at 00:53
@RNar Ah, but thats calling from a `pd.DataFrame` object, while this is from a `.csv` file, which are vastly different situations. — Zizouz212, Nov 10 '15 at 00:54
@Zizouz212 a `.csv` file which he is parsing as a pd.DataFrame with `pandas.read_csv` — R Nar, Nov 10 '15 at 00:56
@RNar Exactly, the questions are different, and therefore, the answers will be different as well. They need to be nearly identical to be considered duplicates. — Zizouz212, Nov 10 '15 at 00:58

score 14 · Accepted Answer · answered Nov 10 '15 at 00:53

14

dt = pandas.read_csv('file.csv', index_col=1, skiprows=1).T.to_dict()

answered Nov 10 '15 at 00:53

Alexander

thanks @Alexander, your soln does not seem to work if any of the values are non-unique – user308827 Nov 10 '15 at 00:57
why do you need the `.T`? – Charlie Parker Feb 20 '17 at 20:52
@Pinocchio It transposes the dataframe so that the index becomes the column names, i.e. the keys to the dictionary. – Alexander Feb 20 '17 at 21:01

Leb · Answer 2 · 2015-11-10T01:00:28.463

5

Duplicating data:

import pandas as pd
from io import StringIO

data="""
A      B
test    23
try     34
"""

df = pd.read_csv(StringIO(data), delimiter='\s+')

Converting to dictioanry:

print(dict(df.values))

Will give:

{'try': 34, 'test': 23}

edited Nov 10 '15 at 01:00

answered Nov 10 '15 at 00:52

Leb

That has nothing to do with what the op is asking, I'm just replicating his data – Leb Nov 10 '15 at 00:58
Then what makes you think this is a suitable answer?!? – Zizouz212 Nov 10 '15 at 00:58
3

All he needs is that `print(dict(df.values))` to get the output he needs. – Leb Nov 10 '15 at 01:00
But how is this reading from a `.csv` file? It's completely wrong. – Zizouz212 Nov 10 '15 at 01:00
He reads the file as a dataframe, then converts to dictionary. I'm not sure how else to put it. – Leb Nov 10 '15 at 01:02
Your answer is the simplest to replicate but a little difficult to understand for rookies. – Jean Paul May 12 '17 at 14:33

2 Answers2