Retrieve Date from DateTime Column in Polars

In polars, to retrieve the date from a datetime column, you can use the dt.date() method. This method extracts just the date (year, month, day) from a datetime column and removes the time portion (hours, minutes, seconds).

Retrieving the date from a datetime column in Polars means extracting only the calendar date (in the format YYYY-MM-DD) from a column that includes both date and time information (such as 2025-05-08 14:30:00). In this article, I will explain retrieve the date from a datetime column in polars.

Key Points –

• Use the dt.date() method to extract the date part (YYYY-MM-DD) from a datetime column.
• The resulting extracted date will be in the date format, not including time (hours, minutes, seconds).
• You can create a new column with the date values while keeping the original datetime column intact.
• The datetime column can be replaced with the date part by using with_columns() and giving the same column name.
• This operation removes the time (hours, minutes, seconds) from the datetime values, providing only the date component.
• After extracting the date, you can filter rows based on specific dates, useful for date-based analysis.
• The extracted date respects the timezone of the datetime column but does not carry time zone info into the new column.
• If the original datetime column contains null values, the result will also contain nulls in the extracted date column.

<a id="h-usage-of-retrieve-date-from-datetime-column"></a>

Usage of Retrieve Date from DateTime Column

To extract the date from a datetime column in Polars, you can use the dt.date() method. This method will extract just the date part (year, month, and day) from the datetime values, discarding the time portion.

To run some examples of retrieving data from the datetime column in polars, let’s create a Polars DataFrame.

import polars as pl
import datetime as dt

df = pl.DataFrame({
    "datetime_col": [dt.datetime(2025, 5, 8, 14, 30)]
})
print("Create DataFrame:\n", df)

Yields below output.

To create a DataFrame and extract the date from a datetime_col in Polars, you can use the dt accessor to access datetime functionalities and retrieve the date part.

# Extract the date part from the datetime column
df2 = df.with_columns(
    pl.col("datetime_col").dt.date().alias("date_only")
)
print("DataFrame with Extracted Date:\n", df2)

Here,

• dt.date() extracts the date part from the datetime values (removes time).
• alias("date_only") names the new column "date_only".

<a id="h-retrieve-date-and-keep-original"></a>

Retrieve Date and Keep Original

To extract the date from a datetime column and keep the original column in the Polars DataFrame, you can use the with_columns() function to add a new column without changing the existing one.

import polars as pl
import datetime as dt

# Create a sample DataFrame
df = pl.DataFrame({
    "datetime_col": [dt.datetime(2025, 5, 8, 14, 30), dt.datetime(2025, 5, 9, 10, 45)]
})

# Add a new column with only the date
df2 = df.with_columns(
    pl.col("datetime_col").dt.date().alias("date_only")
)
print("Original datetime and extracted date:\n", df2)

# Output:
# Original datetime and extracted date:
# shape: (2, 2)
┌─────────────────────┬────────────┐
│ datetime_col        ┆ date_only  │
│ ---                 ┆ ---        │
│ datetime[μs]        ┆ date       │
╞═════════════════════╪════════════╡
│ 2025-05-08 14:30:00 ┆ 2025-05-08 │
│ 2025-05-09 10:45:00 ┆ 2025-05-09 │
└─────────────────────┴────────────┘

Here,

• datetime_col: the original datetime with time
• date_only: the extracted date part (YYYY-MM-DD)

<a id="h-rename-date-column-in-place"></a>

Rename Date Column In-Place

To rename the date column in-place (i.e., replace the original datetime column with just the date), you can apply dt.date() and use alias() with the same column name. Alternatively, to rename the date column while keeping the original datetime column unchanged in Polars, you can use the rename method.

# Extract the date part and rename the column in place
df_with_date = df.with_columns(df["datetime_col"].dt.date().alias("date_only")
).rename({"date_only": "extracted_date"})
print("DataFrame with renamed date column:\n", df_with_date)

# Output:
# DataFrame with renamed date column:
# shape: (2, 2)
┌─────────────────────┬────────────────┐
│ datetime_col        ┆ extracted_date │
│ ---                 ┆ ---            │
│ datetime[μs]        ┆ date           │
╞═════════════════════╪════════════════╡
│ 2025-05-08 14:30:00 ┆ 2025-05-08     │
│ 2025-05-09 10:45:00 ┆ 2025-05-09     │
└─────────────────────┴────────────────┘

Here,

• The date_only column is renamed to extracted_date using the rename() method.
• The original datetime column (datetime_col) is still preserved.

<a id="h-create-a-dataframe-from-multiple-datetimes"></a>

Create a DataFrame From Multiple Datetimes

To extract the date from multiple datetime values in a Polars DataFrame while keeping the original datetime column intact, you can apply dt.date() to the relevant columns.

import polars as pl
import datetime as dt

# Create a DataFrame with multiple datetime values
df = pl.DataFrame({
    "datetime_col": [
        dt.datetime(2025, 1, 1, 10, 0),
        dt.datetime(2025, 2, 14, 18, 45),
        dt.datetime(2025, 5, 8, 9, 15)
    ]
})

# Extract only the date from each datetime
df2 = df.with_columns(
    pl.col("datetime_col").dt.date().alias("date_only")
)
print("DataFrame with multiple datetimes and extracted dates:\n", df2)

# Output:
# DataFrame with multiple datetimes and extracted dates:
# shape: (3, 2)
┌─────────────────────┬────────────┐
│ datetime_col        ┆ date_only  │
│ ---                 ┆ ---        │
│ datetime[μs]        ┆ date       │
╞═════════════════════╪════════════╡
│ 2025-01-01 10:00:00 ┆ 2025-01-01 │
│ 2025-02-14 18:45:00 ┆ 2025-02-14 │
│ 2025-05-08 09:15:00 ┆ 2025-05-08 │
└─────────────────────┴────────────┘

<a id="h-extract-year-month-day-separately"></a>

Extract Year, Month, Day Separately

To extract the year, month, and day separately from a datetime column in a Polars DataFrame, you can use the dt.year(), dt.month(), and dt.day() functions.

import polars as pl
import datetime as dt

# Create a sample DataFrame with datetime values
df = pl.DataFrame({
    "datetime_col": [
        dt.datetime(2025, 5, 8, 14, 30),
        dt.datetime(2025, 6, 15, 10, 45),
        dt.datetime(2026, 7, 20, 9, 15)
    ]
})

# Extract year, month, and day separately
df2 = df.with_columns([
    pl.col("datetime_col").dt.year().alias("year"),
    pl.col("datetime_col").dt.month().alias("month"),
    pl.col("datetime_col").dt.day().alias("day")
])
print("Extracted year, month, and day:\n", df2)

# Output:
# Extracted year, month, and day:
# shape: (3, 4)
┌─────────────────────┬──────┬───────┬─────┐
│ datetime_col        ┆ year ┆ month ┆ day │
│ ---                 ┆ ---  ┆ ---   ┆ --- │
│ datetime[μs]        ┆ i32  ┆ i8    ┆ i8  │
╞═════════════════════╪══════╪═══════╪═════╡
│ 2025-05-08 14:30:00 ┆ 2025 ┆ 5     ┆ 8   │
│ 2025-06-15 10:45:00 ┆ 2025 ┆ 6     ┆ 15  │
│ 2026-07-20 09:15:00 ┆ 2026 ┆ 7     ┆ 20  │
└─────────────────────┴──────┴───────┴─────┘

Here,

• The year, month, and day are extracted into separate columns: year, month, and day.
• The original datetime_col remains intact.

<a id="h-extract-date-then-filter"></a>

Extract Date Then Filter

To extract the date from a datetime column and filter the rows based on that date, you can first use the dt.date() method to extract the date, and then apply a condition to filter the DataFrame according to your criteria.

import polars as pl
import datetime as dt

# Create DataFrame with multiple datetime values
df = pl.DataFrame({
    "datetime_col": [
        dt.datetime(2025, 5, 8, 14, 30),
        dt.datetime(2024, 12, 25, 6, 45),
        dt.datetime(2023, 1, 1, 0, 0),
        dt.datetime(2022, 7, 15, 18, 20)
    ]
})

# Step 1: Extract date and add it as a new column
df_with_date = df.with_columns(
    df["datetime_col"].dt.date().alias("date_only")
)

# Step 2: Filter the DataFrame based on the extracted date
df2 = df_with_date.filter(df_with_date["date_only"] == dt.date(2024, 12, 25))
print("Filtered DataFrame:\n", df2)

# Output:
# Filtered DataFrame:
# shape: (1, 2)
┌─────────────────────┬────────────┐
│ datetime_col        ┆ date_only  │
│ ---                 ┆ ---        │
│ datetime[μs]        ┆ date       │
╞═════════════════════╪════════════╡
│ 2024-12-25 06:45:00 ┆ 2024-12-25 │
└─────────────────────┴────────────┘

Here,

• Extract the Date: We use dt.date() to extract the date part from the datetime_col and create a new column date_only.
• Filter the DataFrame: We filter the rows where date_only matches "2024-12-25".

<a id="h-conclusion"></a>

Conclusion

In conclusion, extracting the date from a datetime column in Polars is simple and efficient using the dt.date() method. Whether you want to create a new column, replace the original, or filter based on date values, Polars offers clear and powerful tools to handle datetime operations with ease.

Happy Learning!!

<a id="h-related-articles"></a>

Related Articles

• Strip Entire Polars DataFrame
• How to Update the Polars DataFrame
• Mapping a Python Dict to a Polars Series
• How to Convert Struct to Series in Polars?
• How to Remove Duplicate Columns in Polars?
• Conditional Assignment in Polars DataFrame
• Check if any Value in a Polars DataFrame is True
• How to Transform a Series of a Polars DataFrame?
• Add a New Column into an Existing Polars DataFrame
• How to Effectively Create Duplicate Rows in Polars?
• Efficient way to Update a Single Element of a Polars DataFrame?
• How to Append a Python List to Another List (Series) of a Polars DataFrame?

<a id="h-reference"></a>

Reference

• https://docs.pola.rs/api/python/stable/reference/expressions/api/polars.datetime_range.html