Opened 13 years ago
Closed 10 years ago
#102 closed defect (fixed)
AZMED reprojection snaps X within 30cm of X origin to origin
Reported by: | ravenAtSafe | Owned by: | Norm Olsen |
---|---|---|---|
Priority: | major | Milestone: | |
Component: | Dictionaries | Version: | |
Keywords: | Cc: |
Description
The value of cs_AnglTest is such that X values within ~30cm of the X origin (center of AZMED/AZMEA system) get snapped to 0 during dynamic reprojections.
CS_azmed.c:
/* Tmp3 and sin_del_lng will both be zero if the point is the origin. Most atan2's (not all) will bomb if both args are zero. */ if (fabs (sin_del_lng) > cs_AnglTest) { Az = atan2 (sin_del_lng,tmp3); sin_Az = sin (Az); cos_Az = cos (Az); } else { /* del_lng is zero, we do a quick and dirty atan2. */ [...]
Change History (2)
comment:1 by , 13 years ago
Owner: | changed from | to
---|
comment:2 by , 10 years ago
Resolution: | → fixed |
---|---|
Status: | new → closed |
Note:
See TracTickets
for help on using tickets.
Changed the threshold for filtering out calls to atan2 when both arguments are dangerously close to zero, such as when converting a point very very close to the origin of the coordinate reference system. New value is 1.0E-12, about 6 microns on the surface of the earth.
Submission produced version 2461.