What is the DataFrame.update function in Polars Python?

Polars, a Rust DataFrame library with Python bindings, excels in high-performance data processing for large datasets. Its parallel processing capabilities and support for various data sources make it an ideal choice for efficient tabular data management with better performance than pandas.

Here, we’ll discuss the update() function of the Polars library.

The update() function

The update() function helps us to merge or align two DataFrames, updating values in the target DataFrame with non-null values from the source DataFrame. This function facilitates the integration of new data into an existing DataFrame, allowing for flexible updating strategies, such as inner, outer, and explicit column-based joins.

Syntax

Here’s the syntax for the update() function:

DataFrame.update(
    other: DataFrame,
    on: str | Sequence[str] | None = None,
    how: Literal['left', 'inner', 'outer'] = 'left',
    *,
    left_on: str | Sequence[str] | None = None,
    right_on: str | Sequence[str] | None = None,
    include_nulls: bool = False,
) → DataFrame[source]
Syntax of update() function

In the syntax above:

  • other passes a DataFrame required to update the original one.

  • on represents the column names that will be joined on.

  • how defines the merging approach: left retains all left rows, inner keeps matching keys, and outer updates existing matches while adding new rows.

  • left_on joins columns of the left DataFrame.

  • right_on joins columns of the right DataFrame.

  • include_nulls states that null values in the right DataFrame will be utilized to update the left DataFrame.

Code example

Let’s discuss a coding example to better understand how this function works: 

import polars as pl
df = pl.DataFrame(
    {
        "EmployeeName": ["John", "Smith", "David", "Ronaldo"],
        "Age": [24, 32, 19, 26],
        "Salary": [100, 300, 250, 320]
    }
)
new_df = pl.DataFrame(
    {
        "Salary": [140, 330, None],
        "LastName": ["Dan", "Rohn", "Michel"],
    }
)
# Simply update DataFrame
print(df.update(new_df))
# Update DataFrame by keeping those rows that are common
print(df.update(new_df, how="inner"))
# Update DataFrame containing all rows in both DataFrames
print(df.update(new_df, how="outer"))
# Explicitly joining columns in each Dataframe, including null values
print(df.update(new_df, left_on="EmployeeName", right_on="LastName", how="outer", include_nulls=True))

Code explanation

Let’s discuss the above code in detail.

  • Line 1: We import the polars library as pl.

  • Lines 3–9: We create a DataFrame named df containing EmployeeName, Age, and Salary columns.

  • Lines 11–16: We create another DataFrame to update the previous one.

  • Line 19: We call the update() function to update the df DataFrame values with the new_df.

  • Line 22: We update DataFrame by keeping those rows that are common by passing how="inner" argument to the function.

  • Line 25: We update DataFrame containing all rows in both DataFrames by passing how="outer" argument to the function.

  • Line 28: We join all columns in each Dataframe, including null values, by passing include_nulls=True argument.

Unlock your potential: Polars in Python series, all in one place!

To continue your exploration of Polars, check out our series of Answers below:

Free Resources

HowDev By Educative. Copyright ©2025 Educative, Inc. All rights reserved