Hi Robert. No, I'm not using one of the maps that ships with SAS but rather ESRI boundary files (which I know are in lat/long) supplied by a third party, an example of which can be found here: http://www.abs.gov.au/ausstats/subscriber.nsf/log?openagent&1270055001_sa4_2011_aust_shape.zip&1270.0.55.001&Data Cubes&B18D49356F3FDA5FCA257801000D6D2E&0&July 2011&23.12.2010&Latest. I am using the EAST and DEGREES options as the longitude values are in degrees and increase to the east. I also don't have the LONGMIN and LONGMAX reversed (SAS gives an error if this is the case). The map produced by PROC GMAP looks fine if I don't clip (I.E. exclude LONGMIN and LONGMAX from the PROC GPROJECT statement) for both the Albers' and gnomonic projections (for some reason Lambert's projection rotates the whole map 180 degrees but that's an issue for another day). And the final map looks fine (apart from the east-west stretch of an unprojected map) if I use the PROJECT=NONE option with clipping. It's when I try to use both that things get interesting: when using Albers' projection the map rotates counterclockwise (see image); when using the gnomonic projection I get a boundary file with just 4 rows with missing X or Y data (and no warning or error in the log)!?! This code... gproject data=sa4_bound out=crop_1 project=albers east degrees; id sa4_code11; run; results in This code... gproject data=sa4_bound out=crop_1 project=albers east degrees longmin=106 longmax=159; id sa4_code11; run; results in Finally if I use no projection, this code gproject data=sa4_bound out=crop_1 project=none east degrees longmin=106 longmax=159; id sa4_code11; run; results in
... View more