Changing data types while reading the file in polars

schema_overrides

Most people read CSV files and then change the data type of columns. That is an amateur way of doing it. The professional approach is to do it all at once. Below is a dataframe showing a Polars function I covered each day in my 100DaysOfPolars.

import polars as pl
pl.read_csv('days_data')

shape: (3, 2)

Function	Day_Count
str	i64
"strip_chars"	1
"pl.nth"	2
"explode"	3

Assign data type on read

To assign a different data type to a column while reading a CSV file, use the option schema_overrides like this:

pl.read_csv('days_data.csv', schema_overrides={'Day_Count': pl.String})

shape: (3, 2)

Function	Day_Count
str	str
"strip_chars"	"1"
"pl.nth"	"2"
"explode"	"3"

There’s more you can learn in my Polars course.