Describe the requested improvement
The old segment-based classification behavior in sits_classify.vector_cube() is deprecated.
When the input is a segs_cube (inherits from vector_cube + raster_cube), sits_classify() now follows the standard raster classification workflow. It applies the temporal model, produces pixel-level probabilities, preserves vector_info, and returns a probs_vector_cube.
Segment-level aggregation is no longer performed in sits_classify().
Associated sits API function
This affects sits_classify(), specifically the S3 method for vector_cube.
Expected behavior, reflecting the actual S3 class:
segs_cube (vector_cube + raster_cube)
-> sits_classify()
probs_vector_cube (derived_vector_cube + segs_cube)
- No new parameter is required.
- The parameter
n_sam_pol, previously used for the old vector classification workflow, emits an explicit deprecation warning:
"Parameter 'n_sam_pol' is deprecated. Segment-level classification is no longer performed by sits_classify(). Segment-based aggregation belongs to sits_label_classification()."
Additional context
sits_classify() remains responsible for temporal probabilistic classification.
- Segment-based class assignment is handled later, during labeling.
- The output preserves
vector_info; the resulting probability cube can still be labelled or analysed using the associated segments.
Describe the requested improvement
The old segment-based classification behavior in
sits_classify.vector_cube()is deprecated.When the input is a
segs_cube(inherits fromvector_cube+raster_cube),sits_classify()now follows the standard raster classification workflow. It applies the temporal model, produces pixel-level probabilities, preservesvector_info, and returns aprobs_vector_cube.Segment-level aggregation is no longer performed in
sits_classify().Associated sits API function
This affects
sits_classify(), specifically the S3 method forvector_cube.Expected behavior, reflecting the actual S3 class:
n_sam_pol, previously used for the old vector classification workflow, emits an explicit deprecation warning:Additional context
sits_classify()remains responsible for temporal probabilistic classification.vector_info; the resulting probability cube can still be labelled or analysed using the associated segments.