How to Create Empty Dataframe With Only Column Names – Definitive Guide

Pandas Dataframe allows you to store data in rows and column format.

You can create empty dataframe with only column names using the pd.DataFrame(columns = column_names_as_list) statement.

Basic Example

import pandas as pd

column_names = ["Col 1", "Col 2", "Col 3"]

df = pd.DataFrame(columns = column_names)

df

• The column names are passed as a list to the columns property.

This tutorial teaches you how to create an empty dataframe with only column names using the pd.Dataframe() constructor.

Create an empty DataFrame with ONLY column names

You can create an empty pandas dataframe with only column names using the pd.Dataframe() constructor.

• The constructor accepts the column names using the columns property.
• Pass column names in an array-like object

Code

The below code demonstrates how to create an empty dataframe with ONLY column names.

• The column names are assigned to the list column_names
• This list is passed to the columns parameter in the dataframe constructor

import pandas as pd

column_names = ["Col 1", "Col 2", "Col 3"]

df = pd.DataFrame(columns = column_names)

df

An empty dataframe will be created with the column names.

Dataframe Will Look Like

	Col 1	Col 2	Col 3

Create an empty DataFrame with column names from another dataframe

To create an empty dataframe with column names from another dataframe,

• Get the column names of another dataframe using the df.columns statement
• Pass it to the columns attribute in the dataframe constructor

Code

df2 = pd.DataFrame(columns = df.columns)

df2

Dataframe Will Look Like

	Col 1	Col 2	Col 3

Create Empty Dataframe With Column names And Column Types

To create an empty dataframe with column names and column types,

• create an empty series with desired dtype
• Assign it to the column names.

If you want to change the column type after creating the dataframe, read How to Change column type in pandas.

Code

The below code demonstrates how to create an empty dataframe with column names and column types.

df = pd.DataFrame({'Col 1': pd.Series(dtype='str'),
                   'Col 2': pd.Series(dtype='int'),
                   'Col 3': pd.Series(dtype='float')})

df.dtypes

You can print the dataframe types using the df.dtypes to see the data type of the columns.

Output

    Col 1     object
    Col 2      int64
    Col 3    float64
    dtype: object

Additional Resources

• How To Combine Two Columns In Pandas
• How to Add Column to Dataframe in Pandas
• How to Rename Column in pandas