Introduction

This repo is prepared to share my forecasting code that predicts the hourly charge load in 2008 based on the historical load data and weather information. There are two sets of data that are used in the training:

• data/load_hist_data.csv: this file records hourly load data from 2005 to 2007
• data/weather_data.csv: this file records hourly temperature data from 2005 to 2007

The final answer is recorded in data/output/probability_estimates.csv

I describe the steps I took in the Jupyter notebook

Installation

1. Clone the repo

   git clone https://github.com/ALS15204/load_prediction.git

2. Build venv: under the repo root

   python3 -m venv ./

3. Install requirements

   pip install -r requirements.txt

(back to top)

Usage

Running Jupyter notebook

1. To use Jupyter note book under the newly built venv

   python -m ipykernel install --user --name=venv

2. Go to the Jupyter notebook to run it

(back to top)

References

I used the following references when building the model

1. US Federal Holidays

(back to top)

License

Distributed under the MIT License. See LICENSE.txt for more information.

(back to top)

Contact

Ronin Wu, PhD - @RoninWu - ronin@ronin-wu.com

Project Link: https://github.com/ALS15204/load_prediction

(back to top)

Acknowledgments

I sincerely thank Mr Anonymous for sharing the data with me. It has been a fun exercise :)