Embedding feature creation into a scikit-learn pipeline

Throughout this chapter, we’ve discussed how to automatically create and select features from time-series data by utilizing tsfresh. Then, we used these features to train a classification model to predict whether an office was occupied at any given hour.

tsfresh includes wrapper classes around its main functions, extract_features and extract_relevant_features, to make the creation and selection of features compatible with the scikit-learn pipeline.

In this recipe, we will set up a scikit-learn pipeline that extracts features from time series using tsfresh and then trains a logistic regression model with those features to predict office occupancy.

How to do it...

Let’s begin by importing the necessary libraries and getting the dataset ready:

1. Let’s import the required libraries and functions:

   import pandas as pd
   from sklearn.pipeline import Pipeline
   from sklearn.linear_model import LogisticRegression...