Skip to content

Simple Testing

To quickly test the tool, follow the steps in the Getting Started section before. You can try it out with the following example inputs:

dataset = ds0              # Sample dataset for testing
forecast_horizon = fh1     # fh1 = 30 minutes ahead
model_name = m6            # m6 = Linear Regression
hyperparameter_no = 'hp1'  # Example hyperparameter set

This is a great way to get familiar with how the tool works before running full-scale tests.

Full Model Testing (All 18 Models)

For a complete evaluation and checking the result against benchmark values, which is available on experiment_result/Archive/Testing Result/testing_benchmark. To test it, follow these steps:

  1. Open the file run_tests.ipynb.
  2. Run all cells without making any changes.
  3. The notebook will automatically: Run all 18 models, compare their outputs against three benchmark results. The results are available in:
    experiment_result/Archive/Testing Result/

⏱ Estimated time: ~1 hour on a personal computer with an Intel i5 processor and 32GB RAM. This tool has undergone full model testing on three different computers, and the result can be seen on 

experiment_result/Archive/Testing Result/20250821_test_result_CEEM Computer.csv
experiment_result/Archive/Testing Result/20250821_test_result_UNSW Laptop.csv
experiment_result/Archive/Testing Result/20250822_test_result_SS Personal Laptop