Python Practice

NumPy and SciPy offer most of the functionality that we need, but we might need a few more libraries.

In this section, we'll use several libraries, which we can quickly install from the terminal, the Jupyter Notebook, or similarly from Anaconda Navigator:

pip install -U tsfresh workalendar astral "featuretools[tsfresh]" sktime

All of these libraries are quite powerful and each of them deserves more than the space we can give to it in this chapter.

Let's start with log and power transformations.

Log and Power Transformations in Practice

Let's create a distribution that's not normal, and let's log-transform it. We'll plot the original and transformed distribution for comparison, and we'll apply a statistical test for normality.

Let's first create the distribution:

from scipy.optimize import minimize
import numpy as np
np.random.seed(0)
pts = 10000
vals = np.random.lognormal(0, 1.0, pts...