Add more options to Mercator projection #1043
Conversation
| def test_easting_northing(): | ||
| false_easting = 1000000 | ||
| false_northing = -2000000 | ||
| crs = ccrs.Mercator(false_easting=false_easting, false_northing=false_northing) |
There was a problem hiding this comment.
E501 line too long (83 > 79 characters)
| false_northing = -2000000 | ||
| crs = ccrs.Mercator(false_easting=false_easting, false_northing=false_northing) | ||
| proj4_str = ('+ellps=WGS84 +proj=merc +lon_0=0.0 +x_0={} +y_0={} ' | ||
| '+units=m +lat_ts=0.0 +no_defs'.format(false_easting, false_northing)) |
There was a problem hiding this comment.
E501 line too long (87 > 79 characters)
|
|
||
|
|
||
| def test_scale_factor(): | ||
| scale_factor = 0.9396926207859084 # Should be same as lat_ts=20 for a sphere |
There was a problem hiding this comment.
E501 line too long (81 > 79 characters)
|
|
||
| def test_scale_factor(): | ||
| scale_factor = 0.9396926207859084 # Should be same as lat_ts=20 for a sphere | ||
| crs = ccrs.Mercator(scale_factor=scale_factor, globe=ccrs.Globe(ellipse='sphere')) |
There was a problem hiding this comment.
E501 line too long (86 > 79 characters)
dopplershift
force-pushed
the
mercator-options
branch
from
ea86c78
to
3bf6349
Compare
lib/cartopy/crs.py
Outdated
| ('units', 'm'), | ||
| ('lat_ts', latitude_true_scale)] | ||
| if scale_factor is not None: | ||
| if latitude_true_scale: |
There was a problem hiding this comment.
You'll have to make the default None now; you can't assume the user didn't pass scale_factor and latitude_true_scale=0 explicitly expecting it to work.
| crs = ccrs.Mercator(scale_factor=scale_factor, | ||
| globe=ccrs.Globe(ellipse='sphere')) | ||
| proj4_str = ('+ellps=sphere +proj=merc +lon_0=0.0 +x_0=0.0 +y_0=0.0 ' | ||
| '+units=m +k_0={} +no_defs'.format(scale_factor)) |
There was a problem hiding this comment.
scale_factor is not formatting consistently on Py 2/3. Might be simpler to just hard-code the result.
dopplershift
force-pushed
the
mercator-options
branch
from
3bf6349
to
f3e3eef
Compare
|
Not sure why circleCI is failing. |
lib/cartopy/crs.py
Outdated
| ('units', 'm')] | ||
|
|
||
| # If it's None, we use the default for Proj4, which is 0.0 |
There was a problem hiding this comment.
The comment reads as if we are going to pass 0.0, but I guess you mean we implicitly use 0.0 because that is what proj4 assumes when we don't use it?
There was a problem hiding this comment.
That's correct. I've tried to clarify that comment.
dopplershift
force-pushed
the
mercator-options
branch
from
f3e3eef
to
1532cc2
Compare
There was a problem hiding this comment.
The joys of mutually exclusive keywords. 😢 (proj's doing, not yours)
I've generally tried to avoid them wherever possible, but I recognise that sometimes one wants to setup their projection based on what they have, and not have to convert it in some way.
Unfortunately, introducing mutually exclusive keywords makes things a real pain for testing equivalence - I believe that is something we have to address (perhaps not in this PR, but generally within cartopy), otherwise the cartopy Projection classes become (a pretty second-rate) direct wrapper on top of proj.
Have you looked into how easy it would be to normalize one into the other by any chance? If that were easy & reliable then that would be my preferred approach to solving this issue (I'd still happily have mutually exclusive keywords, but would instead normalize them in the init).
This adds scale_factor as an additional way to control the scaling (in addition to latitude_true_scale). It also adds support for false_easting and false_northing. These options are all required to be fully capable of representing Mercator projection coming from a CF-compliant netCDF file.
dopplershift
force-pushed
the
mercator-options
branch
from
1532cc2
to
5c7e332
Compare
| proj4_str = ('+ellps=WGS84 +proj=merc +lon_0=0.0 +lat_ts={} ' | ||
| '+units=m +no_defs'.format(lat_ts)) | ||
| proj4_str = ('+ellps=WGS84 +proj=merc +lon_0=0.0 +x_0=0.0 +y_0=0.0 ' | ||
| '+units=m +lat_ts={} +no_defs'.format(lat_ts)) |
There was a problem hiding this comment.
Oh, weird. I don't actually know what is going on here... 😕
Closes #1040.
This adds scale_factor as an additional way to control the scaling (in addition to latitude_true_scale). It also adds support for false_easting and false_northing. These options are all required to be fully capable of representing Mercator projection coming from a CF-compliant netCDF file.