/gudhi-data

Data used in some examples and in gudhi python data fetch

Primary LanguagePythonMIT LicenseMIT

gudhi-data

These datasets are used as examples, for benchmark purposes and in gudhi python data fetch methods.

Do not hesitate to submit a pull request with new model data and sources or to submit an issue of a "wanted" dataset.

License

The root LICENSE file applies to all datasets, except if another license file is located in a dataset directory, then it applies to the datasets in this same directory and subdirectories.

For instance, Stanford bunny LICENSE applies to Stanford bunny point cloud.

Point clouds

Model Name Source Description
Stanford bunny Stanford University OFF file with 35947 points in $\mathbb{R}^3$ and 69451 triangles
Spiral 2D Generated within GUDHI CSV file with 114562 points in $\mathbb{R}^2$
Daily and sports activities UC Irvine ML repository - CC BY 4.0 CSV file with 30000 points in 3d + activity type column
SO3 The SO3 points clouds have been generated with the ISOI software OFF file with 50000 points in $\mathbb{R}^9$

Time Series

Model Name Source Description
Topological Wheels DataShape 10 CSV files where each file has 64 timeseries over 10000 timestamps each, and a $y$ ground truth indicating anomalous regime. See README.