mirror of https://github.com/CGAL/cgal
The LAS format, even binary, is not an exact representation. The coordinates are encoded using and offset and a scaling, for each coordinate. In `write_LAS_with_properties()`, Simon has hard-coded: ``` LASheader header; header.x_scale_factor = 1e-9 * (bbox.xmax() - bbox.xmin()); header.y_scale_factor = 1e-9 * (bbox.ymax() - bbox.ymin()); header.z_scale_factor = 1e-9 * (bbox.zmax() - bbox.zmin()); header.x_offset = bbox.xmin(); header.y_offset = bbox.ymin(); header.z_offset = bbox.zmin(); ``` So, the approximate comparison of coordinates, for the I/O test, should compare using about the same precision. |
||
|---|---|---|
| .. | ||
| doc/Point_set_processing_3 | ||
| examples/Point_set_processing_3 | ||
| include/CGAL | ||
| package_info/Point_set_processing_3 | ||
| test/Point_set_processing_3 | ||