Outside -90/+90 degree range: pygmt.info()/gmt info output for geographic data and -I option in polar latitudes

Hi all

I am generating a set of maps using pygmt.info( data=myGeoDataFrame, spacing=4 ) to set the region. Now, one of the features I am plotting has vertices around -88.8˚S, resulting in a -R-48/128/-92/-76 region definition. As I parse this straight into pygmt.coast() and a Mercator projection, pygmt rightly complains about -92 being outside the -90/+90 degree range (this is for global features, so currently I am not worried too much about the projection).

As the geodataframe has an EPSG:4326 projection, I would have assumed that pygmt.info() is clever enough to just give -90 as southern boundary. The same happens also with OGR_GMT-formatted data ( @Je4326 in header) and gmt info.

Bug or feature request to limit the increment output of (py)gmt info for geographic data to -90/+90 degree range?


/edit: GMT 6.4.0_965fc97-dirty_2021.11.20, pygmt v0.5.0

You are probably not telling GMT that your data are geographic lon/lat. When I use -fg and get a region out of bounds, GMT will say this:

gmt info t.txt -I7 -fg
gmtinfo [WARNING]: Using -I caused wesn[YLO] to become < -90. Reset to -90.


[Note to self: wesn[YLO] really? You could not just write south?]. Of course, we cannot prevent you from using this with a Mercator projection and there is no “default” latitude for that anyway other than < 90. If you plan to do this then you would be better off getting the w e s n values numerically and impose your own limits.

Perhaps the Pygmt wrapper can check for this and pass the -fg flag to the API.

The same happens also with OGR_GMT -formatted data ( @Je4326 in header) and gmt info

That I may be able to fix in the core.

1 Like

Thanks Paul, clear case of RTFM and being too late in the case of gmt info.

PyGMT.info() however doesn’t seem to have a flag for geographic data indeed.

For now, I’ll probably post-process the string from pygmt.info() to make sure that it is -90 < s < 90 like you suggest.


For completeness, I am using for now:

    region = pygmt.info( data=gdf,spacing=4)
    if region[2] < -90:
        region[2] = 89.9.
    if region[3] > 90:
        region[3] = 89.9.

[edited to include @EJFielding’s suggestion]

If you are using the Mercator projection, you will get better results if you make the maximum latitude less than 90 degrees, like 89.9 degrees.

1 Like

CLI GMT master now understands the EPGS4326 in an OGR/GMT file (and hence shape files) at least, but doing this for the dataframe will need to take place in PyGMT.

Thanks @pwessel - Will test with the CLI version!

@EJFielding - much appreciated, have adopted my script & post accordingly.