Python/Pandas convert CSV into dictionary and write to new file

Question

I have a CSV with cluters and their latitude/longitudes like this:

cluster_label,latitude,longitude
0,39.18193382,-77.51885109
1,39.18,-77.27
2,39.17917928,-76.6688633
3,39.1782,-77.2617
4,39.1765,-77.1927

I am trying to convert them into a dictionary using Pandas such as this:

{u'0': [39.18193382,-77.51885109],
 u'1': [39.18,-77.27],
 u'2': [39.17917928,-76.6688633],
 u'3': [39.1782,-77.2617],
 u'4': [39.1765,-77.1927],}

I have converted into a dictionary using .T.to_dict() but when I try to write out to a new file I get Attribute Error(s):'dict' object has no attribute 'to_csv/json/etc.'. How can I convert my csv file to a dictionary and then save the resulting dictionary as a new file?

Code

import pandas as pd

data = pd.read_csv('cluster_centroids.csv', delimiter=',', index_col='cluster_label')

dict = data.T.to_dict()
dict.to_csv('centroidDict.csv')

Aklys · Accepted Answer · 2017-04-16 21:34:10Z

1

If you are trying to output a dataframe to csv you can do the following rather than convert something to a dictionary.

import pandas as pd

file = 'test.csv'

df = pd.read_csv('test.csv', index_col='cluster_label')

df['co-ords'] = tuple(zip(df.latitude, df.longitude))
df.drop(['latitude', 'longitude'],axis=1, inplace=True)

df.to_csv('centroidDict.csv')

but if you want to just have the dictionary written to a file you may want to take a look at use pickle to retain the information for python to use later.

answered Apr 16, 2017 at 21:34

Aklys

4811 gold badge5 silver badges15 bronze badges

Sign up to request clarification or add additional context in comments.

3 Comments

andrewr Over a year ago

Yes, I would like to retain the dictionary structure in a new file. The purpose is to then read this dictionary file as a positions in a networkx graph as mentioned in the 2nd answer here: stackoverflow.com/questions/19915266/…

Aklys Over a year ago

Well there is a number of options for writing the dictionary to a file. You can pickle it if you don't need it to be human readable when stored. You could us JSON or XML to store it as readable. You can always keep the data in the DataFrame above and convert it at the time you need to. But it looks like NetworkX may be able to take information from a DataFrame as well so I'd potentially look at doing that. But I'm not familiar with NetworkX so unsure in that regards.

andrewr Over a year ago

Okay, thanks for your help. I attempted to just convert to a dict and have NetworkX read the dict within the script instead of another file but ran into errors there as well. I'll have to tinker some and ask more questions on the topic.

Collectives™ on Stack Overflow

Python/Pandas convert CSV into dictionary and write to new file

1 Answer 1

3 Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

3 Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related