Introduction
This is the implementation of the model proposed in a proceeding paper of AAAI 2021, A Multi-step-ahead Markov Conditional Forward Model with Cube Perturbations for Extreme Weather Forecasting.
torch==1.4.0
torchfile==0.1.0
torchtext==0.4.0
torchvision==0.5.0
numpy==1.17.2
tqdm==4.42.0
sklearn==0.23.1
jupyter==1.0.0
jupyter-client==5.3.4
jupyter-console==6.0.0
jupyter-core==4.6.0
The ExtremeWeather dataset can be downloaded from here
Use data_preprocess_optimal_area.ipynb
to preprocess:
- In the step "2 Read Data", change the path of the data downloaded from previous step.
- Change the
label_type
in the step "3. Get labels of original image ready for optimal area and time finding" to your target extreme weather event. - Set the path of the output folder in the step "7. Saving Preprocessed Data"
- Then, go through all the preprocessing steps for the experiments input.
Use cnn_exp.py
to conduct experiment:
- Change the
folder_name
andvalidation_folder_name
in thecnn_exp.py
to your preprocessed data. - Modifiy the output path at the end of the code. Or you can directly use the default output path.
python3 cnn_exp.py {perturbation rate} {perturbation type} {cube size}