Revision history [back]

To change the data types of columns within a DataFrame using Polars, you can use the cast method.

Here's an example:

Assume you have a DataFrame df with columns col1 and col2 of data types int64 and float64, respectively, and you want to change col2 to a int64 data type.

import polars as pl

# Create a sample DataFrame
df = pl.DataFrame({
    'col1': [1, 2, 3],
    'col2': [1.0, 2.0, 3.0]
})

# Change col2 to int64 data type
df = df.lazy()
df = df.select([
    pl.col('col1'),
    pl.col('col2').cast(pl.Int64)
])
df = df.collect()

print(df)

Output:

shape: (3, 2)
┌───────┬───────┐
│ col1  ┆ col2  ┆
│ ---   ┆ ---   ┆
│ i64   ┆ i64   ┆
╞═══════╪═══════╡
│ 1     ┆ 1     ┆
│ 2     ┆ 2     ┆
│ 3     ┆ 3     ┆
└───────┴───────┘

Here, we first use the lazy method to create a lazy DataFrame to avoid loading the data into memory. Then, we select the two columns, and use the cast method to change the data type of col2 to int64. Finally, we call the collect method to execute the computation and get the result as a DataFrame.