spherical_conv#
- hypercoil.functional.sphere.spherical_conv(data: Tensor, coor: Tensor, scale: float = 1, r: float = 1, max_bin: int = 10000, truncate: float | None = None)[source]#
Convolve data on a 2-sphere with an isotropic Gaussian kernel.
This is implemented in pretty much the dumbest possible way, but it works. Here is a likely more efficient method that requires Lie groups or some such thing: https://openreview.net/pdf?id=Hkbd5xZRb
See
spatial_conv()for implementation details.- Dimension:
- data : \((*, C, N)\)
* denotes any number of intervening dimensions, C denotes the number of data channels, and N denotes the number of data observations per channel.
- coor : \((*, N, D)\)
D denotes the dimension of the space in which the data are embedded.
Output : \((*, C, N)\)
- Parameters:
- dataTensor
Tensor containing data observations, which might be arrayed into channels.
- coorTensor
Tensor containing the spatial coordinates associated with each observation in data.
- scalefloat (default 1)
Scale parameter of the Gaussian kernel.
- rfloat (default 1)
Radius of the sphere.
- max_binint
Maximum number of points to include in a distance computation. If you run out of memory, try decreasing this.
- truncatefloat or None (default None)
Maximum distance at which data points can be convolved together.
- Returns:
- data_convTensor
The input data convolved with the kernel. Each channel is convolved completely separately as of now.