This container extracts TDA features using the Giotto-TDA Python package. A vector describing each dimension (x, y, z) is output:
- Persistence entropy [0, 1, 2]
- Number of points [0, 1, 2]
- Amplitude
- Landscape [0, 1, 2]
- Bottleneck [0, 1, 2]
- Wasserstein [0, 1, 2]
- Betti [0, 1, 2]
- Silhouette [0, 1, 2]
- Heat [0, 1, 2]
- Persistence Image [0, 1, 2]
The units of traits are as follow:
- Min_x, min_y, min_z, max_x, max_y, and max_z are UTM coordinates, so they are in units of meters
- Num_points is the total count of points in the point cloud
- Hull volume, oriented bounding box, and axis aligned bounding box are in units of cubic meters
- The rest are topological values (those listed above) and are unitless.
- -p, --pointclouds | Input point clouds either single or list
- -c, --cpu | Number of CPUs to use for multiprocessing. It is recommended to be 0.5*(total available CPUs)
- -o, --outdir | Output directory for CSV file containing feature values
- -f, --filename | Output filename for CSV file containing feature values