Electricity Demand Forecasting on the CAISO grid

This repository contains the code, models, and data utilized in the Electricity Demand Forecasting capstone project as presented by Anton Bosneaga, Arsh Hothi, Alan Jian, Naman Patel, and Aska Zaman. For a summary of the project, take a look at the project website and interact with the demo website.

Project Directory Organization

• EDA: Contains ipython notebooks pertaining to exploratory data analysis
• data-collection: Contains ipython notebooks that make API calls to download data necessary for the project
• data-files: Contain downloaded data files in .csv format
• deploy: Contain files that aid in the deployment of the demo website
• model_notebooks: Contain ipython notebooks and R notebooks that help train and tune models used in the project
• models: Contain pickled models ready for deployment
• predictions: Contain .csv files with predicted electricity demand corresponding to each particular time period

Project Dependencies

You can access this project's specific dependencies in the deploy/requirements.txt file. To download the required dependencies, we recommend setting up a new conda environment, and running pip install -r deploy/requirements.txt.