Convert a pandas dataframe to dictionary with one column as key and other column as multiple values

user_16521450 0
Viewed 3725

I want to convert this dataframe into a dictionary where for one single label as key, I store multiple tweets as value. Can someone help?

enter image description here

pandas machine-learning nlp numpy python
2 Answers

Assuming your data frame is variable name is "df" then below may help :

temp = df.groupby(['labels']).apply(lambda x: x['tweets'].tolist()).to_dict()
print(temp)
user_60058760

To get your expected result you can run e.g.:

result = df.groupby('labels')['tweets'].apply(list).to_dict()

Details:

  1. df.groupby('labels') - groups source rows.
  2. ['tweets'] - takes only tweets column (from each group).
  3. apply(list) - converts tweets from the current group into a list. You don't even need to use any explicit lambda function. So far (the result of groupby and apply) is a pandasonic Series.
  4. to_dict() - converts this Series to a dictionary.

For your source data (shortened a bit) the result is:

{'EXP': ['if you missed', 'the emotional'],
 'QUE': ['the neverending'],
 'STM': ['katie couric', 'a yearold nigerian']}
user_73884770
Related