Python Feature Engineering Cookbook

There's more...

In this recipe, we identified the local maxima and minima in a time series for a single household. But how could we determine the local maxima and minima for several houses? We can combine the use of pandas groupby(), pandas agg(), and user-defined functions to achieve this goal. To demonstrate how to do this, we will use the toy customer transactions dataset from Featuretools. Let's get started:

For more details about the customer transactions dataset from Featuretools, visit the Getting ready section of the Aggregating transactions with mathematical operations recipe in this chapter.

1. Let's import the required libraries and data:

import numpy as npimport pandas as pdimport matplotlib.pyplot as pltfrom scipy.signal import ...