**Cook’s distance** is used to identify influential observations in a regression model.

The formula for Cook’s distance is:

**D _{i}** = (r

_{i}

^{2}/ p*MSE) * (h

_{ii}/ (1-h

_{ii})

^{2})

where:

**r**_{i }is the i^{th}residual**p**is the number of coefficients in the regression model**MSE**is the mean squared error**h**_{ii}is the i^{th}leverage value

Essentially Cook’s distance measures how much all of the fitted values in the model change when the i^{th} observation is deleted.

The larger the value for Cook’s distance, the more influential a given observation.

A general rule of thumb is that any observation with a Cook’s distance greater than 4/n (where *n* = total observations) is considered to be highly influential.

This tutorial provides a step-by-step example of how to calculate Cook’s distance for a given regression model in Python.

**Step 1: Enter the Data**

First, we’ll create a small dataset to work with in Python:

import pandas as pd #create dataset df = pd.DataFrame({'x': [8, 12, 12, 13, 14, 16, 17, 22, 24, 26, 29, 30], 'y': [41, 42, 39, 37, 35, 39, 45, 46, 39, 49, 55, 57]})

**Step 2: Fit the Regression Model**

Next, we’ll fit a simple linear regression model:

**import statsmodels.api as sm
#define response variable
y = df['y']
#define explanatory variable
x = df['x']
#add constant to predictor variables
x = sm.add_constant(x)
#fit linear regression model
model = sm.OLS(y, x).fit() **

**Step 3: Calculate Cook’s Distance**

Next, we’ll calculate Cook’s distance for each observation in the model:

#suppress scientific notation import numpy as np np.set_printoptions(suppress=True) #create instance of influence influence = model.get_influence() #obtain Cook's distance for each observation cooks = influence.cooks_distance #display Cook's distances print(cooks) (array([0.368, 0.061, 0.001, 0.028, 0.105, 0.022, 0.017, 0. , 0.343, 0. , 0.15 , 0.349]), array([0.701, 0.941, 0.999, 0.973, 0.901, 0.979, 0.983, 1. , 0.718, 1. , 0.863, 0.713]))

By default, the **cooks_distance()** function displays an array of values for Cook’s distance for each observation followed by an array of corresponding p-values.

For example:

- Cook’s distance for observation #1:
**.368**(p-value: .701) - Cook’s distance for observation #2:
**.061**(p-value: .941) - Cook’s distance for observation #3:
**.001**(p-value: .999)

And so on.

**Step 4: Visualize Cook’s Distances**

Lastly, we can create a scatterplot to visualize the values for the predictor variable vs. Cook’s distance for each observation:

import matplotlib.pyplot as plt plt.scatter(df.x, cooks[0]) plt.xlabel('x') plt.ylabel('Cooks Distance') plt.show()

**Closing Thoughts**

It’s important to note that Cook’s Distance should be used as a way to *identify* potentially influential observations. Just because an observation is influential doesn’t necessarily mean that it should be deleted from the dataset.

First, you should verify that the observation isn’t a result of a data entry error or some other odd occurrence. If it turns out to be a legit value, you can then decide if it’s appropriate to delete it, leave it be, or simply replace it with an alternative value like the median.