Solar Irradiance - IFT6759

This project was created as part of the UdeM course IFT6759 (https://admission.umontreal.ca/cours-et-horaires/cours/IFT-6759/). The objective of this project is to predict present and future (up to 6 hours) solar irradiance at any point on a map of continental united states by only using present and past remote sensing readings from satellites (GOES-13). We propose a machine learning model for prediction of solar irradiance. Refer to the report and presentation included in this reporistory for more details.

Team 08

• Alexander Peplowski
• Harmanpreet Singh
• Marc-Antoine Provost
• Mohammed Loukili

To run the evaluation script:

1. cd scripts/
2. Update submit_evalution.sh 
3. sbatch submit_evalution.sh

OR

1. cd scripts/
2. Update run_evaluatior.sh 
3. Run run_evaluatior.sh

K-Fold Strategy

• Hold out 1 year of data
• No use of k-fold until pipeline is optimized

Coding Standards

• Lint your code as per PEP8 before submitting a pull request
• Pull requests are required for merging to master for major changes
• Use your own branch for major work, don't use master
• No large files allowed in git
• Mark task in progress on Kanban before starting work

To setup a new local environment:

module load python/3.7
virtualenv ../local_env
source ../local_env/bin/activate
pip install -r requirements_local.txt

To setup a new server node environment:

module load python/3.7
virtualenv ../server_env --no-download
source ../server_env/bin/activate
pip install --no-index -r requirements.txt

OR, if no requirement.txt file is available:

pip install --no-index tensorflow-gpu==2 pandas numpy tqdm

To evaluate results from server locally using tensorboard:

Run the commands to synchronize data from the server and to launch tensorboard:

./rsync_data.sh
./run_tensorboard.sh

Use a web browser to visit: http://localhost:6006/