You're reading from Time Series Analysis with Python Cookbook Practical recipes for exploratory data analysis, data preparation, forecasting, and model evaluation

Product type Paperback

Published in Apr 2025

Publisher

ISBN-13 9781805124283

Length 98 pages

Edition 2nd Edition

Languages

Python

Concepts

Data Analysis

Author (1):

Tarek A. Atwan

View More author details

Table of Contents (13) Chapters

1. Time Series Analysis with Python Cookbook, Second Edition: Practical recipes for exploratory data analysis, data preparation, forecasting, and model evaluation FREE CHAPTER

2. Getting Started with Time Series Analysis

3. Reading Time Series Data from Files

4. Reading Time Series Data from Databases

5. Persisting Time Series Data to Files

6. Persisting Time Series Data to Databases

7. Working with Date and Time in Python

8. Handling Missing Data

9. Outlier Detection Using Statistical Methods

10. Exploratory Data Analysis and Diagnosis

11. Building Univariate Time Series Models Using Statistical Methods

12. Additional Statistical Modeling Techniques for Time Series

13. Outlier Detection Using Unsupervised Machine Learning

Forecasting univariate time series data with ARIMA

In this recipe, you will explore ARIMA, using the statsmodels package for implementation. ARIMA stands for Autoregressive Integrated Moving Average, which combines three principal components: the autoregressive or AR(p) model, the moving average or MA(q) model, and an integrated process or I(d) which applies differencing to the data.

An ARIMA model is characterized by the p, d, and q parameters, an ARIMA model for a non-seasonal time series is described by the notation ARIMA(p, d, q). The p and q parameters denote the orders or lags; for example, an AR of order p and MA of order q. They are called lags as they represent the number of “past” periods we need to consider. You may also come across another reference for p and q, namely polynomial degrees.

ARIMA models can handle non-stationary time series data through differencing, a time series transformation technique, to make a non-stationary time series stationary. The integration...