How to Convert Index of a Pandas Dataframe Into a Column

Asad Riaz Feb 02, 2024 Pandas Pandas DataFrame
  1. df.index to Add Index as a New Column
  2. reset_index With rename_axis to Rename the Current Index Column Name
  3. Use the set_index Method to Convert Column to Index
  4. MultiIndex to Set Multiple Layers of indexes on column
How to Convert Index of a Pandas Dataframe Into a Column

We will introduce various methods to convert the index of a Pandas DataFrame into a column, like df.index, set_index, and reset_index with rename_axis to rename the index.

We will also introduce how we can apply Multi-Index to a given DataFrame with multiple layers of indexes.

df.index to Add Index as a New Column

The simplest way to add index as the column is by adding df.index as a new column to DataFrame.

Example Codes:

# python 3.x
import pandas as pd

df = pd.DataFrame(
    [(1, 2, None), (None, 4, None), (5, None, 7), (5, None, None)],
    columns=["a", "b", "d"],
)
df["index"] = df.index
print(df)

Output:

     a    b    d  index1
0  1.0  2.0  NaN       0
1  NaN  4.0  NaN       1
2  5.0  NaN  7.0       2
3  5.0  NaN  NaN       3

reset_index With rename_axis to Rename the Current Index Column Name

We can change the name of our index, then use reset_index to a series:

# python 3.x
import pandas as pd

df = pd.DataFrame(
    [(1, 2, None), (None, 4, None), (5, None, 7), (5, None, None)],
    columns=["a", "b", "d"],
)
df = df.rename_axis("index").reset_index()
print(df)

Output:

   index    a    b    d
0      0  1.0  2.0  NaN
1      1  NaN  4.0  NaN
2      2  5.0  NaN  7.0
3      3  5.0  NaN  NaN

Use the set_index Method to Convert Column to Index

We can convert any column to index using the set_index method.

# python 3.x
import pandas as pd

df = pd.DataFrame(
    [(1, 2, None), (None, 4, None), (5, 4, 7), (5, 5, None)], columns=["a", "b", "d"]
)
df.set_index("b", inplace=True)
print(df)

Output:

     a    d
b          
2  1.0  NaN
4  NaN  NaN
4  5.0  7.0
5  5.0  NaN

Or if we want to remove the index name, as in the original, we can do df.index.name = None:

# python 3.x
import pandas as pd

df = pd.DataFrame(
    [(1, 2, None), (None, 4, None), (5, 4, 7), (5, 5, None)], columns=["a", "b", "d"]
)
df.set_index("b", inplace=True)
df.index.name = None
print(df)

Output:

     a    d
2  1.0  NaN
4  NaN  NaN
4  5.0  7.0
5  5.0  NaN

MultiIndex to Set Multiple Layers of indexes on column

We can use MultiIndex.from_product() function to make a MultiIndex as follow:

# python 3.x
import pandas as pd
import numpy as np

index = pd.MultiIndex.from_product(
    [["Burger", "Steak", "Sandwich"], ["Half", "Full"]], names=["Item", "Type"]
)
df = pd.DataFrame(
    index=index, data=np.random.randint(0, 10, (6, 4)), columns=list("abcd")
)
print(df)

Output:

               a  b  c  d
Item     Type            
Burger   Half  0  3  9  1
         Full  2  2  0  5
Steak    Half  8  4  5  5
         Full  5  8  0  7
Sandwich Half  2  8  9  5
         Full  4  4  5  9
Enjoying our tutorials? Subscribe to DelftStack on YouTube to support us in creating more high-quality video guides. Subscribe

Related Article - Pandas DataFrame