scRL.GridCore.project_back

scRL.GridCore.project_back(gres, key, neighbors=15, w=None, negative=False)[source]

This function projects grid-based data back to the original points. It uses a Gaussian kernel to weight the influence of the nearest grids on each point, optionally adjusted by a user-defined weighting factor. The results can be scaled to positive values and are stored back in the original data structure.

Parameters:

  • gres – Grids results

  • key – Key in the grids dict which must be a continuous variable

  • neighbors – Nearest grid number to be considered (Default: 15)

  • w – Annotation to weight the projected pseudotime at cell level (Default: None)

  • negative – Whether the value is non zero or not (Default: False)

Return type:

None