This repository contains the implementation of the EAM model proposed in the paper titled "Help, I Shrunk My Savings! Assessing the Carbon Reduction Potential for Video Streaming from Short-Term Coding Changes" by Daniel Schien, which has been submitted to ICIP 2023. The EAM model is built on top of the eam-core library which is available at https://github.com/sust-cs-uob/eam-core.
groupings.yml: This file contains the groupings used in the model to group time periods t_00 to t_23.iplayer_views_per_hour.xlsx: This file contains the hourly views data for iPlayer programs.short_term_model.xlsx: This file contains the input data required for running the short-term model.short-term-model.yml: This file contains the configuration for the short-term model used in the EAM model.results/: This directory will be created automatically when the model is run and will contain the output results.summary_v3.xlsx: This file contains the summary of the output of the model.
To use the EAM model, you need to have Python 3 installed on your system. You can install the required dependencies, including the eam-core library, by running the following commands in your terminal
Python 3:
sudo apt update
sudo apt install python3
python3 --version
sudo apt upgrade python3
eam-core:
git clone https://github.com/sust-cs-uob/eam-core.git
pip install -e .
ICIP2023 Model:
git clone https://github.com/sust-cs-uob/ICIP2023-EAM-model.git
To run the EAM model, execute the following command in your terminal:
eam-core -a ci -l -c ci -sd -a ci <path to short-term-model.yml>
NOTE: Depending on your directory layout you might need to change the paths within the short-term-model.yml
The output results will be saved in the results/raw/ directory.
For citation information, please see the CITATION.cff file.