Feature/move tests to integration

[NicolasColombi]

Changes proposed in this PR:

Move lazy unittest which loads data multiple time to the integration folder.

List of tests moved:

• climada/util/test/test_lines_polys_handler/test_calc_geom_impact_polys (~ 15.5 sec)
• climada/hazard/test/test_storm_europe/test_icon_read (~ 4.5 sec)
• climada/hazard/test/test_storm_europe/test_from_footprints (~ 6 sec)
• climada/hazard/test/test_ibtracs_with_basin (~6 sec)
• climada/hazard/test/test_cutoff_tracks (~5 sec)
• climada/hazard/test/test_base/test_write_read_pass (4.5 sec)
• climada/hazard/test/test_base/test_raster_to_vector_pass (3.5 sec)
• climada/hazard/test/test_base/test_reproject_raster_pass (3.6 sec)

This PR decrease the cost of unittest by around 51s. All tests are placed in the integration folder either under test_hazard.py under a new classes, or in the new file test_util.py.

The above tests share in common the following characteristics:

• they all load data (multiple times for the most of them)
• almost all of them call other climada functions than the one they are testing

PR Author Checklist

• Read the Contribution Guide
• Correct target branch selected (if unsure, select develop)
• Descriptive pull request title added
• Source branch up-to-date with target branch
• Documentation updated
• Tests updated
• Tests passing
• No new linter issues
• Changelog updated

PR Reviewer Checklist

• Read the Contribution Guide
• CLIMADA Reviewer Checklist passed
• Tests passing
• No new linter issues

[peanutfun]

Excellent initiative, @NicolasColombi, thanks a lot! 🙌

I used black to format the new file and the inserted code in test_hazard.py. I also cut the duplicated code from test_util.py by simply importing the stuff from climada.util.test.test_lines_polys_handler. This seems to work fine.

On my system, one test of test_hazard.py fails. This might be related to #678? As Jenkins does not run integration tests on pull requests, I do not know if the test would fail there as well. @NicolasColombi, does the test run through at your end?

$ python climada/test/test_hazard.py

======================================================================
ERROR: test_icon_read (__main__.TestStormEurope)
test reading from icon grib
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/ldr.riedel/coding/climada_python/climada/test/test_hazard.py", line 184, in test_icon_read
    haz = StormEurope.from_icon_grib(
  File "/Users/ldr.riedel/coding/climada_python/climada/hazard/storm_europe.py", line 506, in from_icon_grib
    intensity=sparse.csr_matrix(stacked.gust.T),
  File "/Users/ldr.riedel/mambaforge/envs/climada_env_3.9/lib/python3.9/site-packages/xarray/core/common.py", line 276, in __getattr__
    raise AttributeError(
AttributeError: 'Dataset' object has no attribute 'gust'

[NicolasColombi]

Hi @peanutfun, Thanks a lot for the review!! 😄 I run all the tests again and they pass, including test_icon_read. I guess you are right, it must be related to #678... Do we want to ask @chahank to run the test on his machine to see if it also fails ? Alternatively, we wait that #678 get fixed and then you can try again from yours @peanutfun :)