How to Select Last Column of Polars DataFrame

In Polars, selecting the last column of a DataFrame means retrieving the column that appears at the final position (rightmost side) of the given DataFrame. This is helpful when the column names are unknown in advance or when you need to access the last column dynamically. It is particularly useful for datasets where column positions may change, allowing you to retrieve the most recently added or last column without specifying its name explicitly. In this article, I will demonstrate how to select the last column of a polars DataFrame.

Key Points–

• Use bracket notation with df.columns[-1] to select the last column.
• Use df.columns[-1] to access the name of the last column by indexing the column list.
• Convert the last column to a Polars Series using the to_series() method.
• Use get_column() to retrieve a column by its name, including the last column.
• Use select() with df.columns[-1] to select the last column.
• Add the last column to the DataFrame using with_columns() and last() functions.
• Use select() with pl.col() to select the last column dynamically by referring to its name.
• Using df.columns[-1] is a lightweight operation and performs well in selecting the last column.

<a id="h-usage-of-select-last-column-of-polars-dataframe"></a>

Usage of Select Last Column of Polars DataFrame

Selecting the last column of a Polars DataFrame is an essential operation in data manipulation, especially when you want to focus on the last attribute or when the number of columns is dynamic and you cannot rely on knowing the exact column name.

To run some examples of how to select the last column of a polars DataFrame, let’s create a Polars DataFrame.

import polars as pl

technologies= {
    'Courses':["Spark","PySpark","Hadoop","Python","Pandas"],
    'Fees' :[22000,25000,23000,24000,26000],
    'Duration':['30days','50days','35days', '40days','55days'],
    'Discount':[1000,2300,1000,1200,2500]
          }
df = pl.DataFrame(technologies)
print("Original DataFrame:\n", df)

Yields below output.

To select the last column of the DataFrame using select() with pl.col(), you can specify the column dynamically by using df.columns[-1] in combination with pl.col() function.

# Selecting the last column dynamically
df2 = df.select(pl.col(df.columns[-1]))
print("Last Column:\n", df2)

Here,

• df.columns[-1] dynamically retrieves the name of the last column ('Discount' in this case).
• pl.col(df.columns[-1]) tells Polars to select that column.
• df.select(pl.col(…)) returns a DataFrame containing just the last column.

<a id="h-select-last-column-using-bracket-notation"></a>

Select Last Column Using Bracket Notation []

You can select the last column of a Polars DataFrame using bracket notation ([]). This method retrieves the last column as a Polars Series. To do so, you can access the column dynamically by its name, using df.columns[-1].

# Selecting the last column 
# Using bracket notation
df2 = df[df.columns[-1]]
print("Last Column (Series):\n", df2)

# Output;
# Last Column (Series):
# shape: (5,)
# Series: 'Discount' [i64]
[
	1000
	2300
	1000
	1200
	2500
]

Here,

• df.columns[-1] dynamically retrieves the name of the last column ('Discount' in this case).
• df[df.columns[-1]] selects that column from the DataFrame. This returns a Polars Series instead of a DataFrame.

This will give you the last column as a Polars Series. If you need it as a DataFrame, you can use the to_frame() method

# Using to_frame() method
df2 = df[df.columns[-1]].to_frame()
print("Last Column:\n", df2)

# Output:
# Last Column:
# shape: (5, 1)
┌──────────┐
│ Discount │
│ ---      │
│ i64      │
╞══════════╡
│ 1000     │
│ 2300     │
│ 1000     │
│ 1200     │
│ 2500     │
└──────────┘

<a id="h-select-last-column-using-get-column"></a>

Select Last Column Using get_column()

To select the last column of a Polars DataFrame using the get_column() method, pass the column name (which can be dynamically accessed with df.columns[-1]). This function returns the specified column as a Polars Series.

# Selecting the last column using get_column()
df2 = df.get_column(df.columns[-1])
print("Last Column (Series):\n", df2)

# Output:
# Last Column (Series):
# shape: (5,)
# Series: 'Discount' [i64]
[
	1000
	2300
	1000
	1200
	2500
]

Here,

• df.columns[-1] dynamically retrieves the name of the last column ('Discount').
• df.get_column(df.columns[-1]) selects the column by name and returns it as a Polars Series.

<a id="h-select-last-column-using-select-with-df-columns"></a>

Select Last Column Using select() with df.columns

You can use the select() method directly with df.columns to select the last column of a Polars DataFrame. This approach lets you specify the column name using df.columns[-1]. By referencing the column name through df.columns[-1], you can select the last column without needing to use pl.col().

# Selecting the last column directly 
# Using df.columns[-1]
df2 = df.select(df.columns[-1])
print("Last Column (DataFrame):\n", df2)

# Output:
# Last Column (DataFrame):
# shape: (5, 1)
┌──────────┐
│ Discount │
│ ---      │
│ i64      │
╞══════════╡
│ 1000     │
│ 2300     │
│ 1000     │
│ 1200     │
│ 2500     │
└──────────┘

Here,

• df.columns[-1] gets the name of the last column (in this case, 'Discount').
• df.select(df.columns[-1]) selects the last column, returning it as a DataFrame.

<a id="h-selecting-the-last-column-using-with-columns-and-last"></a>

Selecting the Last Column Using with_columns() and last()

To select the last column of a Polars DataFrame using with_columns() and last(), you can use the with_columns() method, but instead of adding or modifying columns, you can use it in combination with last() to reference the last column dynamically.

# Use .with_columns() and .last() to select the last column
df2 = df.with_columns(pl.col(df.columns[-1]).alias("last_column"))
print("Last Column (DataFrame):\n", df2)

# Output:
# Last Column (DataFrame):
# shape: (5, 5)
┌─────────┬───────┬──────────┬──────────┬─────────────┐
│ Courses ┆ Fees  ┆ Duration ┆ Discount ┆ last_column │
│ ---     ┆ ---   ┆ ---      ┆ ---      ┆ ---         │
│ str     ┆ i64   ┆ str      ┆ i64      ┆ i64         │
╞═════════╪═══════╪══════════╪══════════╪═════════════╡
│ Spark   ┆ 22000 ┆ 30days   ┆ 1000     ┆ 1000        │
│ PySpark ┆ 25000 ┆ 50days   ┆ 2300     ┆ 2300        │
│ Hadoop  ┆ 23000 ┆ 35days   ┆ 1000     ┆ 1000        │
│ Python  ┆ 24000 ┆ 40days   ┆ 1200     ┆ 1200        │
│ Pandas  ┆ 26000 ┆ 55days   ┆ 2500     ┆ 2500        │
└─────────┴───────┴──────────┴──────────┴─────────────┘

Here,

• The with_columns() method is primarily used to add new columns to a DataFrame.
• pl.col(df.columns[-1]) dynamically selects the last column by referencing df.columns[-1] and using pl.col() to access it.
• alias("last_column") renames the new column to "last_column", making the output clearer and more meaningful.

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

Conclusion

In conclusion, selecting the last column of a Polars DataFrame can be done in various ways, depending on your preference for readability and functionality. Whether you choose bracket notation [], the to_series() method, get_column(), or select(), each approach efficiently allows you to dynamically access the last column.

Happy Learning!!

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

Related Articles

• Polars Filter by Column Value
• Convert Polars String to Integer
• Select Polars Columns by Index
• Polars Rename Columns to Lowercase
• Convert DataFrame to Series in Polars
• Polars Sum DataFrame Columns With Examples
• Reorder Columns in a Specific Order Using Polars
• How to Convert a Polars DataFrame to Python List?
• How to Add a Column with Numerical Value in Polars

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

References

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