ENH: use pyproj.CRS when possible

[snowman2]

Needs tests for new version & probably needs current tests fixed.
Much to do here, but it is a start.

[snowman2]

Likely will address (but remains to be seen):
#943
#245
#65

[snowman2]

Lots of pyproj version checking going on. Would you prefer to just pin pyproj>=2.2.0?

[codecov]

Codecov Report

Merging #998 into master will decrease coverage by 1.82%.
The diff coverage is 70.21%.

@@            Coverage Diff             @@
##           master     #998      +/-   ##
==========================================
- Coverage   89.22%   87.39%   -1.83%     
==========================================
  Files          18       18              
  Lines        1642     1674      +32     
==========================================
- Hits         1465     1463       -2     
- Misses        177      211      +34

Impacted Files	Coverage Δ
geopandas/geoseries.py	93.33% <60%> (-5.96%)	⬇️
geopandas/io/file.py	83.65% <72.72%> (-11.09%)	⬇️
geopandas/tools/crs.py	82.35% <77.77%> (-2.84%)	⬇️
geopandas/base.py	93.96% <85.71%> (-0.31%)	⬇️
geopandas/io/sql.py	78.57% <0%> (-17.86%)	⬇️
geopandas/geodataframe.py	96.84% <0%> (-0.91%)	⬇️
geopandas/plotting.py	91.73% <0%> (-0.87%)	⬇️
geopandas/tools/overlay.py	55.72% <0%> (-0.53%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 01f96f6...19d335c. Read the comment docs.

[codecov]

Codecov Report

Merging #998 into master will decrease coverage by 1.85%.
The diff coverage is 65.45%.

@@            Coverage Diff             @@
##           master     #998      +/-   ##
==========================================
- Coverage   89.72%   87.86%   -1.86%     
==========================================
  Files          20       20              
  Lines        1936     1978      +42     
==========================================
+ Hits         1737     1738       +1     
- Misses        199      240      +41

Impacted Files	Coverage Δ
geopandas/tools/sjoin.py	96.92% <100%> (ø)	⬆️
geopandas/base.py	90.62% <54.54%> (-3.93%)	⬇️
geopandas/geoseries.py	89.71% <62.5%> (-4.3%)	⬇️
geopandas/tools/crs.py	82.35% <77.77%> (-2.84%)	⬇️
geopandas/io/file.py	90% <85.71%> (-4.74%)	⬇️
geopandas/io/sql.py	82.14% <0%> (-14.29%)	⬇️
geopandas/_compat.py	85.71% <0%> (-14.29%)	⬇️
geopandas/tools/_show_versions.py	88.31% <0%> (-2.6%)	⬇️
geopandas/plotting.py	91.17% <0%> (-2.11%)	⬇️
... and 2 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 29add0a...1b58545. Read the comment docs.

[snowman2]

One issue I see is that PROJ does not consider epsg:4326 and +init=epsg:4326 to be equal, whereas fiona does:

>>> crs_wkt = 'GEOGCS["WGS 84",DATUM["WGS_1984",SPHEROID["WGS 84",6378137,298.257223563,AUTHORITY["EPSG","7030"]],AUTHORITY["EPSG","6326"]],PRIMEM["Greenwich",0,AUTHORITY["EPSG","8901"]],UNIT["degree",0.0174532925199433,AUTHORITY["EPSG","9122"]],AUTHORITY["EPSG","4326"]]'
>>> cc1 = CRS(crs_wkt)
>>> cc1
<Geographic 2D CRS: EPSG:4326>
Name: WGS 84
Axis Info [ellipsoidal]:
- Lat[north]: Geodetic latitude (degree)
- Lon[east]: Geodetic longitude (degree)
Area of Use:
- undefined
Datum: World Geodetic System 1984
- Ellipsoid: WGS 84
- Prime Meridian: Greenwich

>>> cc2 = CRS("+init=epsg:4326 +type=crs")
>>> cc2
<Geographic 2D CRS: +init=epsg:4326 +type=crs>
Name: WGS 84
Axis Info [ellipsoidal]:
- lon[east]: Longitude (degree)
- lat[north]: Latitude (degree)
Area of Use:
- name: World
- bounds: (-180.0, -90.0, 180.0, 90.0)
Datum: World Geodetic System 1984
- Ellipsoid: WGS 84
- Prime Meridian: Greenwich

The difference is subtle, but it is mostly the difference between the axis being geodetic or not.
This is an issue when writing to a file and reading in the projection back in.

[jorisvandenbossche]

@snowman2 Thanks a lot for starting this!

I opened #1003 for some of the more general discussion points. Will try to give more detailed code feedback here later.

[snowman2]

Not sure where it is coming from but I don't see epsg:26018 on any google search or on:
http://epsg.io/?q=26018.
And it seems PROJ does not recognize it either and is causing tests to fail.

[snowman2]

Using pyproj 2.2.0 (current master):

[snowman2]

I came up with a version that returns a dictionary. Only one test on Travis failed due to precision on python 2. The changes are on a separate commit and can be easily removed depending on the direction you would like to proceed in. If you like the dict option, I can fix that python 2 test to require less precision when testing the difference.

[snowman2]

Closing for preference to #1101