Pandas DataFrame.describe: Generate Descriptive Statistics

Pandas DataFrame.describe

The DataFrame.describe method in pandas is used to generate descriptive statistics of a DataFrame. This includes measures such as mean, standard deviation, min, max, and percentiles for numerical columns or count, unique, top, and frequency for object columns.

Syntax

The syntax for DataFrame.describe is:

DataFrame.describe(percentiles=None, include=None, exclude=None)

Here, DataFrame refers to the pandas DataFrame on which the descriptive statistics are generated.

Parameters

ParameterDescription
percentilesA list of numbers between 0 and 1 to specify the percentiles to include. Defaults to [.25, .5, .75] if None.
includeSpecifies the data types to include. Use 'all' to include all data types, or specify a list of data types (e.g., ["float", "int"]). Defaults to None (only numeric columns).
excludeSpecifies the data types to exclude. Cannot be used together with include.

Returns

A DataFrame containing descriptive statistics. For numeric data, this includes mean, standard deviation, min, max, and percentiles. For object data, it includes count, unique, top, and frequency.

Examples

Descriptive Statistics of a DataFrame

This example demonstrates how to use describe to generate descriptive statistics for numerical columns in a DataFrame.

Python Program

import pandas as pd

# Create a DataFrame
df = pd.DataFrame({
    'A': [1, 2, 3, 4, 5],
    'B': [10, 20, 30, 40, 50],
    'C': ['a', 'b', 'a', 'b', 'a']
})

# Generate descriptive statistics
result = df.describe()
print(result)

Output

         A     B
count  5.0   5.0
mean   3.0  30.0
std    1.58 15.81
min    1.0  10.0
25%    2.0  20.0
50%    3.0  30.0
75%    4.0  40.0
max    5.0  50.0

Descriptive Statistics Including All Data Types

This example demonstrates how to include all data types in the descriptive statistics using include='all'.

Python Program

import pandas as pd

# Create a DataFrame
df = pd.DataFrame({
    'A': [1, 2, 3, 4, 5],
    'B': [10, 20, 30, 40, 50],
    'C': ['a', 'b', 'a', 'b', 'a']
})

# Generate descriptive statistics for all data types
result = df.describe(include='all')
print(result)

Output

          A     B    C
count   5.0   5.0    5
unique  NaN   NaN    2
top     NaN   NaN    a
freq    NaN   NaN    3
mean    3.0  30.0  NaN
std     1.58 15.81  NaN
min     1.0  10.0  NaN
25%     2.0  20.0  NaN
50%     3.0  30.0  NaN
75%     4.0  40.0  NaN
max     5.0  50.0  NaN

Custom Percentiles for Descriptive Statistics

This example demonstrates how to use the percentiles parameter to include custom percentiles in the output.

Python Program

import pandas as pd

# Create a DataFrame
df = pd.DataFrame({
    'A': [1, 2, 3, 4, 5],
    'B': [10, 20, 30, 40, 50]
})

# Generate descriptive statistics with custom percentiles
result = df.describe(percentiles=[.1, .9])
print(result)

Output

          A     B
count   5.0   5.0
mean    3.0  30.0
std     1.58 15.81
min     1.0  10.0
10%     1.4  14.0
50%     3.0  30.0
90%     4.6  46.0
max     5.0  50.0

Summary

In this tutorial, we explored the DataFrame.describe method in pandas. Key takeaways include:

  • Using describe to compute descriptive statistics for numerical columns by default.
  • Including all data types with the include='all' parameter.
  • Customizing the output by specifying percentiles with the percentiles parameter.
  • Excluding specific data types using the exclude parameter.

Python Libraries