metadata
license: cc-by-nc-sa-4.0
This dataset contains six families of datasets as npz files that can be read directly as follows:
import numpy as np
ds = np.read('/path/to/dataset/*.npz')["Y"]
Within each family we have the train and test datasets already splitted up for easy use,
train_X, train_Y, test_X, test_Y
e.g. Use d1X_train_pad_flat.npz, d1Y_train_pad_flat.npz, d1X_test_pad_flat.npz and d1Y_test_pad_flat.npz as inputs to your SciML models. Our sample code is available in our github repo.
#Detailed Description of Each Family of datasets
- d1 : One input time step and one output time step
- d2 : Three input time steps and one output time step
- d3 : Five input time steps and one output time step
- d4 : Three input time steps and three output time steps
- d5 : Five input time steps and three output time steps
- d6 : Eight input time steps and three output time steps
For full description of these datasets, please refer to our website : https://lobster-app-cbyg9.ondigitalocean.app/