Pandas to_numeric is not downcasting integer column

Question

I have a dataframe with a column with dtype('int64'). The values in the column range from 0-10. The dataframe has 770K rows and 56 columns of different types. When I run the code below, I get dtype('int64'). I would have thought that the result would have been at a minimum to downcast to int32 or int16. Here's a replicable example.

import pandas as pd

df = pd.DataFrame([x for x in range(10)]*77000, columns=['recommendation'])
df.dtypes
df.recommendation.apply(lambda x: pd.to_numeric(x, downcast='integer')).dtypes

try pd.to_numeric(df.recommendation,downcast='integer').dtypes — Ben.T
– Ben.T, Commented Oct 26, 2018 at 13:51

sds · Accepted Answer · 2018-10-26 13:59:35Z

6

The apply method works cell-by-cell, so it cannot figure out that the whole column can be downcast. You need to call to_numeric on the whole column, as indicated by Ben in comment:

pd.to_numeric(df.recommendation,downcast='integer')

answered Oct 26, 2018 at 13:59

sds

60.6k31 gold badges178 silver badges304 bronze badges

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

Collectives™ on Stack Overflow

Pandas to_numeric is not downcasting integer column

1 Answer 1

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Related