landscape-evolution-ml data set and models
收藏资源简介:
This contains the data set and models generated by the landscape-evolution-ml code hosted in https://github.com/ricsatjr/landscape-evolution-ml. This public repository was created from refactored versions of the original codes which are stored in a private github repository. elevts* files: samples of steady state elevation grids generated by the original internal code (corresponding to the current public repository's pipeline/01_generate_landscapes.py at commit f82e60f). Sample elevts files of both the original and refactored codes were compared and were found to be the same. features-all-abdda36.pkl: pickle file containing dataframe of features and labels for the 950 landscapes; generated from the current public repository's pipeline/02_extract_features.py at commit abdda36 using the rasnet files created by the original code as input (sample rasnet files of both the original and refactored codes were compared and were found to be the same). nested-cv-results-full-u-kh-ks-00f0686.pkl: pickle file containing model training results (including the models) with individual parameters (U, Kh, Ks) as labels; generated by pipeline/03_train_models.py at commit 00f0686 using features-all-abdda36.pkl as input. nested-cv-results-full-u_ks-kh_ks-00f0686.pkl: pickle file containing model training results (including the models) with parameter ratios (U/Ks, Kh/Ks) as labels; generated by pipeline/03_train_models.py at commit 00f0686 using features-all-abdda36.pkl as input.



