This repository contains Python code I wrote for forecasting 1-D time series using a simple ARIMA model.
The code is not optimized for performance in any way, but it is useful for experimenting and data exploration.
The segmenting algorithms use NumPy's least squares fitting routine, so naturally it depends on NumPy. The script requires matplotlib to display the plots. It requires pandas for reading and writing of csv-style data. It also requires sklearn for the RMSE metric used
You can run the code to see example output by running the simple_forecaster.py script.
The example uses shampoo sales data (included) and also the UCL ML Household Power Consumption data (must be downloaded and put into the household_power_consumption folder)