search

datalab-dev / quintessence_analysis

All the scripts we use for analysis
0 stars 0 forks source link

In compute subsets of embeddings use groups instead of indices #52

Closed avkoehl closed 3 years ago

avkoehl commented 3 years ago

df.groupby("Date").groups -> returns dictionary group -> group labels df.groupby("Date").indices -> returns dictionary group -> group indices

These may not be the same thing in our data if the _id isn't ordered exactly like indices!

Also double check the list_group_by function, make sure to return labels not indices!

avkoehl commented 3 years ago

list_group_by from quintessence.nlp works Screen Shot 2021-02-19 at 2 07 00 PM

avkoehl commented 3 years ago

somehow it looks like I actually did this right. Amazing