Fix bugs and use real data in the plotting docs #251

adybbroe · 2020-02-06T11:30:10Z

Fix bugs and use real data in the doc pages but hide code needed for doc testing

There was also one duplicated section which is now fixed.

The idea is that the user should be able more or less to generate the same plots by going from top to bottom in a notebook or ipython prompt.

Signed-off-by: Adam.Dybbroe [email protected]

…doctesting Signed-off-by: Adam.Dybbroe <[email protected]>

Signed-off-by: Adam.Dybbroe <[email protected]>

adybbroe · 2020-02-06T13:28:23Z

Should be ready now.
One thing which is lacking is some help on how to get the background blue marble ready for Cartopy. Didn't find a nice description/manual for this online when I was looking now. But locally I have this json file:

➜  raster more images.json 
{"__comment__": "JSON file specifying the image to use for a given type/name and resolution. Read in by cartopy.mpl.geoaxes.read_user_background_images.",
  "BM": {
    "__comment__": "Blue Marble Next Generation, July ",
    "__source__": "https://neo.sci.gsfc.nasa.gov/view.php?datasetId=BlueMarbleNG-TB",
    "__projection__": "PlateCarree",
    "low": "BM.jpeg",
    "high": "BMNG_hirez.png"},
    "pop": {
    "__comment__": "Population density from Nasa earth observations website",
    "__source__": "https://neo.sci.gsfc.nasa.gov/view.php?datasetId=SEDAC_POP",
    "__projection__": "PlateCarree",
    "high": "population.jpg"}
}

and a jpeg file BM.jpeg in the same directory. And that directory is pointed to by CARTOPY_USER_BACKGROUNDS

I actually have these four files in my local directory:

BM.jpeg
BMNG_hirez.png
images.json
population.jpg

Probably only BM.jpeg is needed
Any suggestions welcome!

djhoese · 2020-02-06T14:39:06Z

docs/source/plot.rst

- >>> tb37v = np.arange(1000)
- >>> area_def = load_area('areas.yaml', 'pc_world')
- >>> swath_def = SwathDefinition(lons, lats)
+ >>> area_def = load_area('areas.yaml', 'pc_world') # doctest: +SKIP


Isn't this area load just overwriting the area_def created in the lines above?

Yes, right. But the stuff above is hidden, and does not show up on RTD. Therefore the additional line for RTD. The area is not defined and on Travis we don't have access to areas.yaml and certainly not in the home dir.

Ah, pyresample doesn't have a builtin areas 🤦‍♂️ Got it.

djhoese

Thanks for looking at this Adam. I'm a little confused by the doctest skips. Does doctest only run the code in .. doctest sections or do it run all code that it finds? Can we just not doctest some of these parts? It seems like some lines have a skipped doctest and others don't, but shouldn't they work? Like the load_area calls should be loading the areas.yaml from the package, right?

Signed-off-by: Adam.Dybbroe <[email protected]>

adybbroe · 2020-02-06T15:11:35Z

I am also not pleased with it really. I am using this doctest-hide to help doctesting but hiding it for the user.
I haven't found a way to skip a whole block when testing.

djhoese · 2020-02-20T14:21:46Z

Looks like there are some import issues on the tests. @adybbroe Could you look at these?

Signed-off-by: Adam.Dybbroe <[email protected]>

coveralls · 2020-04-22T20:37:06Z

Coverage increased (+0.05%) to 91.969% when pulling 9e70cd5 on adybbroe:fix-documentation-error-cartopy-plotting into 3d6d38d on pytroll:master.

codecov · 2020-04-22T20:44:07Z

Codecov Report

Merging #251 into master will increase coverage by 0.04%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master     #251      +/-   ##
==========================================
+ Coverage   91.94%   91.99%   +0.04%     
==========================================
  Files          42       42              
  Lines        8098     8181      +83     
==========================================
+ Hits         7446     7526      +80     
- Misses        652      655       +3

Impacted Files	Coverage Δ
pyresample/plot.py	`42.37% <ø> (+0.70%)`	⬆️
pyresample/utils/_proj4.py	`86.95% <0.00%> (-1.57%)`	⬇️
pyresample/test/test_geometry.py	`97.75% <0.00%> (-0.15%)`	⬇️
pyresample/bucket/__init__.py	`98.91% <0.00%> (-0.06%)`	⬇️
pyresample/test/test_grid.py	`100.00% <0.00%> (ø)`
pyresample/test/test_bucket.py	`100.00% <0.00%> (ø)`
pyresample/test/test_kd_tree.py	`97.28% <0.00%> (+0.05%)`	⬆️
pyresample/test/test_utils.py	`97.37% <0.00%> (+0.13%)`	⬆️
pyresample/geometry.py	`82.55% <0.00%> (+0.14%)`	⬆️
pyresample/kd_tree.py	`92.47% <0.00%> (+0.17%)`	⬆️
... 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 3d6d38d...9e70cd5. Read the comment docs.

Signed-off-by: Adam.Dybbroe <[email protected]>

adybbroe · 2020-05-04T21:34:14Z

Closing and reopening the PR to try trigger a RTD build

adybbroe · 2020-05-04T21:42:15Z

I think I am ready with this PR now, and now one can see the doc-pages being build on RTD for PRs.
Except that I here see the # doctest: +SKIP (I suppose they will be gone once merged?) I believe it is ready. Perhaps @TomLav @mraspaud @djhoese or someone else want to go through it quickly?

djhoese · 2020-05-07T20:28:26Z

@mraspaud @adybbroe I think this should be good now. I made one important change to the code that could cause issues for some people: I remove the matplotlib.use line in the save_quicklook function. It was using a warn keyword argument that is being deprecated in matplotlib 3.3 and I personally don't think a function should be setting the matplotlib backend for the user. If this was a known problem and was needed in the past we can revisit it. Otherwise I'm going to assume modern matplotlib behaves better for this.

Fix bugs and use real data in the doc pages but hide code needed for …

2b7c526

…doctesting Signed-off-by: Adam.Dybbroe <[email protected]>

adybbroe self-assigned this Feb 6, 2020

adybbroe requested review from mraspaud and djhoese February 6, 2020 11:30

adybbroe added bug documentation labels Feb 6, 2020

Skip doc testing the resampling, it is done already earlier

551c25c

Signed-off-by: Adam.Dybbroe <[email protected]>

adybbroe marked this pull request as ready for review February 6, 2020 13:28

djhoese reviewed Feb 6, 2020

View reviewed changes

Make Travis happy

b0fe478

Signed-off-by: Adam.Dybbroe <[email protected]>

djhoese mentioned this pull request Feb 20, 2020

Switch to pytest for CI and remove Python <3.4 support #254

Merged

Adam.Dybbroe added 2 commits April 22, 2020 17:28

Merge branch 'master' into fix-documentation-error-cartopy-plotting

c4c6958

Try make Travis happy with the doc tests

2ad5904

Signed-off-by: Adam.Dybbroe <[email protected]>

Adam.Dybbroe added 4 commits May 4, 2020 16:01

Add back code example so it is visible in the doc pages

b5e4597

Signed-off-by: Adam.Dybbroe <[email protected]>

Divide plottting example pages into three source files

b1240bb

Signed-off-by: Adam.Dybbroe <[email protected]>

Improve and condense the plot examples pages

7e1e1a0

Signed-off-by: Adam.Dybbroe <[email protected]>

Rename main plot rst file back to original naming

7c26f10

Signed-off-by: Adam.Dybbroe <[email protected]>

adybbroe closed this May 4, 2020

adybbroe reopened this May 4, 2020

adybbroe closed this May 4, 2020

adybbroe reopened this May 4, 2020

djhoese added 2 commits May 7, 2020 15:24

Rewrite plotting docs to render correctly

c1d0cad

Remove extra whitespace

f3db969

Skip more sphinx doctests when matplotlib isn't installed

06d7b40

djhoese added 5 commits May 8, 2020 09:40

Fix indentation of plot_projections example

1677410

Debugging weird travis Windows reading behavior

2c52125

Fix symbolic link reading in travis windows

aebcfcd

Try using functions that actually exist

0f6ba62

Skip doctest downloading cartopy shapefiles

9e70cd5

djhoese merged commit 92ad021 into pytroll:master May 12, 2020

djhoese changed the title ~~Fix bugs and use real data in the doc pages but hide code needed for …~~ Fix bugs and use real data in the plotting docs May 12, 2020

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Fix bugs and use real data in the plotting docs #251

Fix bugs and use real data in the plotting docs #251

adybbroe commented Feb 6, 2020 •

edited

Loading

adybbroe commented Feb 6, 2020 •

edited

Loading

djhoese Feb 6, 2020

adybbroe Feb 6, 2020

djhoese Feb 6, 2020

djhoese left a comment

adybbroe commented Feb 6, 2020

djhoese commented Feb 20, 2020

coveralls commented Apr 22, 2020 •

edited

Loading

codecov bot commented Apr 22, 2020 •

edited

Loading

adybbroe commented May 4, 2020

adybbroe commented May 4, 2020

djhoese commented May 7, 2020

Fix bugs and use real data in the plotting docs #251

Fix bugs and use real data in the plotting docs #251

Conversation

adybbroe commented Feb 6, 2020 • edited Loading

adybbroe commented Feb 6, 2020 • edited Loading

djhoese Feb 6, 2020

Choose a reason for hiding this comment

adybbroe Feb 6, 2020

Choose a reason for hiding this comment

djhoese Feb 6, 2020

Choose a reason for hiding this comment

djhoese left a comment

Choose a reason for hiding this comment

adybbroe commented Feb 6, 2020

djhoese commented Feb 20, 2020

coveralls commented Apr 22, 2020 • edited Loading

codecov bot commented Apr 22, 2020 • edited Loading

Codecov Report

adybbroe commented May 4, 2020

adybbroe commented May 4, 2020

djhoese commented May 7, 2020

adybbroe commented Feb 6, 2020 •

edited

Loading

adybbroe commented Feb 6, 2020 •

edited

Loading

coveralls commented Apr 22, 2020 •

edited

Loading

codecov bot commented Apr 22, 2020 •

edited

Loading