It is possible now to apply lower resolution to mapping
of float-integer-float and decrease size of mapped file.