If the data has a validity mask as well as a map, the mapped data is also masked. When mapping a value datum that has a mask value of zero (invalid), the mapping process will first look for an entry in the map that corresponds to the "invalid" value. If one does not exist (for example, the value is NaN), the mapped data is set to the pad value. In either case, the data remains masked. The mask is resized to correspond to the size of the mapped value data to properly mask it. For example, the value data may have the dimensions of
width = 100, height = 50, elements = depth = time = 1.
If the map dimensions are
map width = 3, map height = whatever, map element = map depth = map time = 1,
the resulting value (and mask) size will be
width = 100, height = 50, elements = 3, depth = time = 1.
All other data, such as location, are assumed to be unaffected, and are transferred to the output object.
kmapdata -i image:lizard -o lizard.mappedThis will map the value data from "image:mandril" through its map, resulting in an output object with of dimensionality Width = 512, Height = 480, Depth = 1, Time = 1, Elements = 3, and no map.