I'm struggling to understand how the geo projections work in practice.

The info a the CRS is as follows: {'ellps': 'WGS84', 'no_defs': True, 'proj': 'longlat'}

Does it mean that the coordinates are planar? if so,

  • how do I know what it was projected from (I need areas preserved, so this is important)
  • do I need to do something to displayed it on a map?

My region is South Africa- - how do I choose the best coordinate system?

  • All shapefiles coming form the gov sources seem planar. Is it a good/best practice to distribute files in this format? – Anna Magdalena Feb 21 '16 at 9:21
  • If you have a .prj file with your .shp collection than open it with notepad. There should be information about CRS. The choice may depends on you or the requirement of customer data. Shapefile is a good (probably universal) format to distribute data, a lot of programs can read it. – Artec Feb 21 '16 at 10:18
  • There was no .prj in the distribution, unfortunately. – Anna Magdalena Feb 22 '16 at 0:12

The CRS you give is a set of proj4 parameters and values:


Your one is just saying the associated numbers are lat-long on the WGS84 spheroid. If that's all you've got then you've no way of knowing if the data originally came from a planar coordinate system.

If you want to convert them to a planar coordinate system, then just find the coordinate code for the UTM zone that covers your region.

| improve this answer | |
  • I have to say that those parameters are some software's interpretation/internal representation of, probably, PROJ.4 parameters. The linked page doesn't list 'longlat' for one thing. – mkennedy Feb 21 '16 at 12:57
  • No, but it tells you that if you type "proj -l" you'll see the valid values for the "+proj" parameter, including lonlat : Lat/long (Geodetic) – Spacedman Feb 21 '16 at 13:00
  • @Spacedman you thought of proj -lp, but that does not include longlat. Furthermore, proj +proj=longlatfails as well as using lonlat and latlon. – AndreJ Feb 21 '16 at 13:09
  • Agh, blinded to the 'g' in longlat. Weird, suspect OP has an odd typo somewhere. – Spacedman Feb 21 '16 at 16:27
  • @Spacedman, "if the data originally came from a planar coordinate system." ? Isn't it that oryginally the data always comes from a certain spheroid and is projected onto sth (e.g. plane)? I have no problems understanding transformations and coord. systems (I'm come from alg. geom.), yet when it comes to geo-tasks, the original form of data is not clear to me. Shouldn't all data come as: points/shapes on a spheroid X were ..., and after a projection given by Y (e.g. UTM with parameters..), it ends up as ...? Am I missing sth? – Anna Magdalena Feb 22 '16 at 0:22

Not the answer you're looking for? Browse other questions tagged or ask your own question.