Deep learning for the occurrence of tipping points: training data

This data accompanies the manuscript by Chengzuo Zhuge et al. “Deep learning for the occurrence of tipping points” and the Github repository https://github.com/zhugchzo/dl_occurrence_tipping. It contains the model time series data that are used to train the deep learning algorithm. The directory increased_bifurcation contains 150k time series (50k Fold, Hopf, Transcritical respectively) with parameter increasing and the directory decreased_bifurcation contains 150k time series (50k Fold, Hopf, Transcritical respectively) with parameter decreasing. The directory pitchfork contains 100k time series (50k supercritical and subcritical pitchfork respectively) with parameter increasing. Both directories contain files labels.csv and groups.csv which provide numbers corresponding to the labels (The tipping points) and groups (Training, Validation, Test) for each time series respectively.