Use Pandas to Calculate Statistics in Python

Performing various complex statistical operations in python can be easily reduced to single line commands using pandas. We will discuss some of the most useful and common statistical operations in this post. We will be using the Titanic survival dataset to demonstrate such operations.

# Import Pandas Library
import pandas as pd

# Load Titanic Dataset as Dataframe
dataset = pd.read_csv('train.csv')

# Show dataset
# head() bydefault show 
# 5 rows of the dataframe
dataset.head()

Output:

1. Mean:

Calculates the mean or average value by using DataFrame/Series.mean() method.

Syntax: DataFrame/Series.mean(self, axis=None, skipna=None, level=None, numeric_only=None, **kwargs)

Parameters:

• axis: {index (0), columns (1)}

          Specify the axis for the function to be applied on.

• skipna:  This parameter takes bool value, default value is True

           It excludes null values when computing the result.

• level: This parameter takes int value or level name, default value is None.

          If the axis is a MultiIndex, count along a particular level, collapsing into a Series.

• numeric_only: This parameter takes bool value, default value is None

           Include only float, int, boolean columns. If None, will attempt to use everything, then use only numeric  data values. Not implemented for Series.

• **kwargs: Additional arguments to be passed to the function.

Returns:  Mean of Series or DataFrame (if level specified)

Code:

# Calculate the Mean 
# of 'Age' column
mean = dataset['Age'].mean()

# Print mean
print(mean)

Output: 

29.69911764705882

2. Median:

Calculates the median value by using DataFrame/Series.median() method.

Syntax: DataFrame/Series.median(self, axis=None, skipna=None, level=None, numeric_only=None, **kwargs)

Parameters:

• axis: {index (0), columns (1)}

          Specify the axis for the function to be applied on.

• skipna:  This parameter takes bool value, default value is True

          It excludes null values when computing the result.

• level: This parameter takes int or level name, default None

          If the axis is a MultiIndex, count along a particular level, collapsing into a Series.

• numeric_only:  This parameter takes bool value, default value is None

          Include only float, int, boolean columns. If value is None, will attempt to use everything, then use only  numeric data.

• **kwargs: Additional arguments to be passed to the function.

Returns:  Median of Series or DataFrame (if level specified)

Code:

# Calculate Median of 'Fare' column
median = dataset['Fare'].median()

# Print median
print(median)

Output: 

14.4542

3. Mode:

Calculates the mode or most frequent value by using DataFrame.mode() method.

Syntax: DataFrame/Series.mode(self, axis=0, numeric_only=False, dropna=True)

Parameters:

• axis: {index (0), columns (1)}

          The axis to iterate over while searching for the mode value:

          0 value or ‘index’ : get mode of each column

          1 value or ‘columns’ : get mode of each row.

• numeric_only:  This parameter takes bool value, default value is False.

           If True, only apply to numeric value columns.

• dropna: This parameter takes bool value, default value is True.

           Don’t consider counts of NaN/None value.

Returns: Highest frequency value. 

Code:

# Calculate Mode of 'Sex' column
mode = dataset['Sex'].mode()

# Print mode
print(mode)

Output: 

0    male
dtype: object

4. Count:

Calculates the count or frequency of non-null values by using DataFrame/Series.count() Method.

Syntax: DataFrame/Series.count(self, axis=0, level=None, numeric_only=False)

Parameters:

• axis: {0 or ‘index’, 1 or ‘columns’}, default value is 0

          If value is 0 or ‘index’ counts are generated for each column. If value is 1 or ‘columns’ counts are                         generated for each row.

• level: (optional)This parameter takes int or str value.

          If the axis is a MultiIndex type, count along a particular level, collapsing into a DataFrame. A str is used   specifies the level name.

• numeric_only:  This parameter takes bool value, default False

          Include only float, int or boolean data.Returns: Return the highest frequency value 

Returns: For each column/row the number of non-null entries. If level is specified returns a DataFrame                              structure.

Code:

# Calculate Count of 'Ticket' column
count = dataset['Ticket'].count()

# Print count
print(count)

Output: 

891

5. Standard Deviation:

Calculates the standard deviation of values by using DataFrame/Series.std() method.

Syntax: DataFrame/Series.std(self, axis=None, skipna=None, level=None, ddof=1, numeric_only=None, **kwargs)

Parameters:

• axis: {index (0), columns (1)}
• skipna: This parameters takes bool value, default value is True.

          Exclude NA/null values. If an entire row/column has NA values, the result will be NA value.

• level: This parameters takes int or level name, default value is None.

          If the axis is a MultiIndex, count along a particular level, collapsing into a Series.

• ddof: This parameter take int value, default value is 1.

          Delta Degrees of Freedom. The divisor used in calculations is N - ddof, where N value represents the  number of elements.

• numeric_only: This parameter takes bool value , default None

          Include only float, int, boolean columns. If None, will attempt to use everything, then use only numeric  data. Not implemented for Series.

Returns: Standard Deviation 

Code:

# Calculate Standard Deviation
# of 'Fare' column
std = dataset['Fare'].std()

# Print standard deviation
print(std)

Output: 

49.693428597180905

6. Max:

Calculates the maximum value using DataFrame/Series.max() method.

Syntax: DataFrame/Series.max(self, axis=None, skipna=None, level=None, numeric_only=None, **kwargs)

Parameters:

• axis: {index (0), columns (1)}

          Specify the axis for the function to be applied on.

• skipna: bool, default True

          It excludes null values when computing the result.

• level: int or level name, default None

          If the axis is a MultiIndex type, count along a particular level, collapsing into a Series.

• numeric_only: bool, default None

           Include only float, int, boolean columns. If None value, will attempt to use everything, then use only  numeric data.

• **kwargs: Additional keyword to be passed to the function.

Returns: Maximum value in Series or DataFrame (if level specified)

Code:

# Calculate Maximum value in 'Age' column
maxValue = dataset['Age'].max()

# Print maxValue
print(maxValue)

Output: 

80.0

7. Min:

Calculates the minimum value using DataFrame/Series.min() method.

Syntax: DataFrame/Series.min(self, axis=None, skipna=None, level=None, numeric_only=None, **kwargs)

Parameters:

• axis: {index (0), columns (1)}

          Specify the axis for the function to be applied on.

• skipna: bool, default True

          It excludes null values when computing the result.

• level: int or level name, default None

          If the axis is a MultiIndex type, count along a particular level, collapsing into a Series.

• numeric_only: bool, default None

           Include only float, int, boolean columns. If None value, will attempt to use everything, then use only  numeric data.

• **kwargs: Additional keyword to be passed to the function.

Returns: Minimum value in Series or DataFrame (if level specified)

Code:

# Calculate Minimum value in 'Fare' column
minValue = dataset['Fare'].min()

# Print minValue
print(minValue)

Output: 

0.0000

8. Describe:

Summarizes general descriptive statistics using DataFrame/Series.describe() method.

Syntax: DataFrame/Series.describe(self: ~ FrameOrSeries, percentiles=None, include=None, exclude=None) 

Parameters:

• percentiles: list-like of numbers, optional
• include: ‘all’, list-like of dtypes or None values (default), optional
• exclude: list-like of dtypes or None values (default), optional,

Returns: Summary statistics of the Series or Dataframe provided.

# Statistical summary
dataset.describe()

Output: