ts-programs-gg 1.0.0

A collection of 7 time series analysis programs

Time Series Programs Library

A Python library containing 7 time series analysis programs for forecasting and analysis using the AirPassengers dataset.

Installation

You can install this package using pip:

pip install .

Or for development mode:

pip install -e .

Requirements

• Python >= 3.7
• pandas >= 1.3.0
• numpy >= 1.21.0
• matplotlib >= 3.4.0
• statsmodels >= 0.13.0
• scikit-learn >= 0.24.0

Usage

After installation, you can import and run any of the 7 programs:

from ts_programs import p1, p2, p3, p4, p5, p6, p7

# Run program 1: Exponential Smoothing Methods
p1()

# Run program 2: Stationarity Testing
p2()

# Run program 3: Moving Averages & ACF/PACF
p3()

# Run program 4: AutoRegressive Models
p4()

# Run program 5: Moving Average Models
p5()

# Run program 6: ARMA Model
p6()

# Run program 7: ARIMA Model
p7()

Programs Overview

p1() - Exponential Smoothing Methods

Compares Simple Moving Average (SMA), Simple Exponential Smoothing (SES), and Holt-Winters forecasting methods. Displays metrics (MAE, MSE, RMSE) and visualization plots.

p2() - Stationarity Testing

Tests for stationarity using Augmented Dickey-Fuller (ADF) and KPSS tests. Compares the original series with white noise.

p3() - Moving Averages & ACF/PACF Plots

Visualizes moving averages (6-month and 12-month) and autocorrelation functions (ACF and PACF).

p4() - AutoRegressive (AR) Models

Fits AR models of various orders (1, 2, 3, 5, 10) and evaluates their performance using MAE, MSE, and RMSE.

p5() - Moving Average (MA) Models

Fits MA models of various orders (1, 2, 3, 5) and evaluates their performance.

p6() - ARMA Model

Fits an ARMA(2,1) model and generates forecasts for the test period.

p7() - ARIMA Model

Fits an ARIMA(2,1,2) model and generates future forecasts.

Data Requirements

All programs expect an AirPassengers.csv file in your current working directory with the following structure:

• Column 1: Month (date column)
• Column 2: #Passengers (numerical values)

You can download the AirPassengers dataset from various sources or use your own time series data with the same format.

License

MIT License