Installation Instructions

1. Clone the repository to your local machine.
2. Create a Python virtual environment.
3. Install the required packages:

pip install -r requirements.txt

⚠️ This may take ~10 minutes. Although newer Python versions may work, the tool was tested on Python 3.12.3.

How to Use The Tool

1. Open the notebook: notebooks/model/run_experiments.ipynb.
2. Fill in the input values for your forecast problem and model specification.
3. Run the notebook.
4. The tool outputs evaluation results to the experiment_result/folder.
5. To evaluate multiple forecast problems and model specifications at once, use notebooks/model/run_experiments_batch.ipynb

Example input values:

dataset = "ds0"            # Dataset for testing
forecast_horizon = "fh1"   # fh1 = 30 minutes ahead
model_name = "m6"          # Linear regression
hyperparameter_no = "hp1"  # Hyperparameter ID

For the list of available datasets & models, how to modify model hyperparameter, how to add a model, how to add a dataset, and exhaustive list of API Reference see the Detailed Guide page.