Pandas is a popular Python library used for data manipulation and analysis. One of its useful functions is the ability to transpose a dataframe. Transposing a dataframe means converting rows into columns and columns into rows, effectively rotating the dataframe by 90 degrees. In this article, we will cover how to transpose a Pandas Dataframe.

What is Transposing?

Transposing a matrix in linear algebra, and consequently in machine learning, entails changing the rows and columns of a matrix. To determine variances and covariances in regression, this operation is frequently used.

Often, a shifted matrix, such as df, is written as df^T.


A matrix’s form can be changed by swapping the rows and columns. Unless the lengths of the rows and columns are the same, in which case the size of the matrix does not change.

Loading a Sample Dataframe

We’ll be utilizing two distinct dataframes for this tutorial. This is due to the fact that the .transpose() function behaves differently based on whether or not your dataframe contains mixed datatypes.

Let’s begin by loading a few test dataframes:

				
					import pandas as pd

# create first dataframe
data1 = {'name': ['Alice', 'Bob', 'Charlie', 'David', 'Emma'],
         'age': [25, 32, 47, 19, 38],
         'gender': ['F', 'M', 'M', 'M', 'F']}
df1 = pd.DataFrame(data1)

# create second dataframe
data2 = {'name': ['Alice', 'David', 'Emma'],
         'salary': [55000, 40000, 82000],
         'city': ['New York', 'Houston', 'Miami']}
df2 = pd.DataFrame(data2)

# print both dataframes
print("DataFrame 1:")
print(df1)

print("\nDataFrame 2:")
print(df2)

Output:

				
					#DataFrame 1:
#       name  age gender
# 0     Alice   25      F
# 1       Bob   32      M
# 2   Charlie   47      M
# 3     David   19      M
# 4      Emma   38      F

# DataFrame 2:
#    name  salary      city
# 0  Alice   55000  New York
# 1  David   40000   Houston
# 2   Emma   82000     Miami

This will create a dataframe with columns for Name, Age, and City, and rows for each of the sample data entries. You can modify the data dictionary to include your own sample data, and pandas will automatically create the dataframe accordingly.

Transposing a Pandas Dataframe

Transposing a Pandas dataframe means to swap its rows and columns, so that the columns become rows and the rows become columns. This is useful for reshaping the data and making it more convenient for certain types of analysis.
In Pandas, you can transpose a dataframe using the transpose() method or the .T attribute.

Here’s an example:

				
					import pandas as pd

# create a sample dataframe
data = {'Name': ['John', 'Mary', 'Peter'],
        'Age': [25, 30, 28],
        'Salary': [50000, 60000, 45000]}
df = pd.DataFrame(data)

# display the original dataframe
print("Original DataFrame:")
print(df)

# transpose the dataframe
transposed_df = df.transpose()

# display the transposed dataframe
print("\nTransposed DataFrame:")
print(transposed_df)

Output:

				
					# Original DataFrame:
#    Name  Age  Salary
# 0   John   25   50000
# 1   Mary   30   60000
# 2  Peter   28   45000

# Transposed DataFrame:
#           0     1      2
# Name    John  Mary  Peter
# Age       25    30     28
# Salary  50000  60000  45000

As you can see, transposing the dataframe switches the rows and columns, and creates a new dataframe where the columns are now rows and the rows are now columns. You can then use this transposed dataframe for further analysis or data manipulation.

Transposing a Pandas Dataframe with Mixed Data Types

Transposing a Pandas dataframe with mixed data types is similar to transposing a regular Pandas dataframe. However, if the original dataframe contains mixed data types, the resulting transposed dataframe will have a data type of object.

Here’s an example:

				
					import pandas as pd

# create a sample dataframe with mixed data types
data = {'Name': ['John', 'Mary', 'Peter'],
        'Age': [25, 30, 28],
        'Salary': [50000, 60000, 45000],
        'Is_Employed': [True, False, True]}
df = pd.DataFrame(data)

# display the original dataframe
print("Original DataFrame:")
print(df)

# transpose the dataframe
transposed_df = df.transpose()

# display the transposed dataframe
print("\nTransposed DataFrame:")
print(transposed_df)

Output:

				
					# Original DataFrame:
#    Name  Age  Salary  Is_Employed
# 0   John   25   50000         True
# 1   Mary   30   60000        False
# 2  Peter   28   45000         True

# Transposed DataFrame:
#                 0      1      2
# Name          John   Mary  Peter
# Age             25     30     28
# Salary       50000  60000  45000
# Is_Employed   True  False   True

As you can see, the resulting transposed dataframe has a data type of object, because it contains mixed data types (i.e. string, integer, boolean).

If you need to perform further analysis on this transposed dataframe, you may need to convert the data types of some columns using the appropriate Pandas methods, such as astype().

Transposing a Dataframe with Missing Values

When transposing a Pandas dataframe with missing values, the missing values will be preserved in the resulting transposed dataframe.

Here’s an example:

				
					import pandas as pd
import numpy as np

# create a sample dataframe with missing values
data = {'Name': ['John', np.nan, 'Peter'],
        'Age': [25, np.nan, 28],
        'Salary': [50000, 60000, np.nan]}
df = pd.DataFrame(data)

# display the original dataframe
print("Original DataFrame:")
print(df)

# transpose the dataframe
transposed_df = df.transpose()

# display the transposed dataframe
print("\nTransposed DataFrame:")
print(transposed_df)

Output:

				
					# Original DataFrame:
#    Name   Age   Salary
# 0   John  25.0  50000.0
# 1    NaN   NaN  60000.0
# 2  Peter  28.0      NaN

# Transposed DataFrame:
#              0        1      2
# Name       John      NaN  Peter
# Age        25.0      NaN   28.0
# Salary  50000.0  60000.0    NaN

As you can see, the missing values in the original dataframe are preserved in the resulting transposed dataframe. If you need to perform further analysis on this transposed dataframe, you may need to handle the missing values using appropriate Pandas methods, such as fillna().

Advantages of Transposing Dataframes

Transposing a Pandas dataframe has several advantages, including:

  • Reshaping the data: Transposing a dataframe can reshape the data to make it more convenient for certain types of analysis. For example, if the original dataframe has observations in rows and variables in columns, transposing the dataframe can change the orientation to have variables in rows and observations in columns.
  • Improving readability: Transposing a dataframe can make the data more readable, especially if there are many variables and few observations. For example, if you have a dataframe with 20 variables and 5 observations, transposing the dataframe can make it easier to read by displaying the variables in rows and the observations in columns.
  • Facilitating data manipulation: Transposing a dataframe can make certain types of data manipulation easier. For example, if you need to group data by variable instead of by observation, transposing the dataframe can make it easier to perform the groupby operation.
  • Simplifying data presentation: Transposing a dataframe can simplify data presentation, especially if you need to present the data in a specific format. For example, if you need to present data in a tabular format with variables in rows and observations in columns, transposing the dataframe can facilitate this presentation.

Overall, transposing a Pandas dataframe can be a useful tool for reshaping, analyzing, and presenting data in a more convenient and readable format.

Wrap up

To learn more about the Pandas .transpose()  function, check out the:
https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.transpose.html


Thanks for reading. Happy coding!