In the cleaner, we implicitly assume that the values in the TIF cannot be negative and explicitly choose -1 as a special value for processing.
In both the cleaner and geopolygonizer, we only use band 1 of a TIF file.
Are these reasonable assumptions? Could a TIF file have negative values? Floats? Are these handled properly? Should we tell the user what is allowed, or perhaps internally handle these different cases?
In the cleaner, we implicitly assume that the values in the TIF cannot be negative and explicitly choose -1 as a special value for processing.
In both the cleaner and geopolygonizer, we only use band 1 of a TIF file.
Are these reasonable assumptions? Could a TIF file have negative values? Floats? Are these handled properly? Should we tell the user what is allowed, or perhaps internally handle these different cases?