Fix severe bug in mask_fillvalues

[schlunma]

Description

This PR fixes a severe bug which lead to a wrong masking of data with mask_fillvalues if the data contains zeros. This has already been discussed 2 years ago and has recently lead to problems.

I also fixed some prospector issues along the way.

Related to #487
Closes #1827

---

Before you get started

• ☝ Create an issue to discuss what you are going to do

Checklist

It is the responsibility of the author to make sure the pull request is ready to review. The icons indicate whether the item will be subject to the 🛠 Technical or 🧪 Scientific review.

• 🧪 The new functionality is relevant and scientifically sound
• 🛠 This pull request has a descriptive title and labels
• 🛠 Code is written according to the code quality guidelines
• 🧪 and 🛠 Documentation is available
• 🛠 Unit tests have been added
• 🛠 Changes are backward compatible
• 🛠 Any changed dependencies have been added or removed correctly
• 🛠 The list of authors is up to date
• 🛠 All checks below this pull request were successful

---

To help with the number pull requests:

• 🙏 We kindly ask you to review two other open pull requests in this repository

[schlunma]

This line lead to the buggy behavior: Since data_hits is interpreted as bool later in the code a 0 in data leads to (undesirable) False in the data and thus to wrong masking.

[codecov]

Codecov Report

Merging #1823 (cd347cb) into main (63267b6) will increase coverage by 0.00%.
The diff coverage is 100.00%.

@@           Coverage Diff           @@
##             main    #1823   +/-   ##
=======================================
  Coverage   91.51%   91.52%           
=======================================
  Files         203      203           
  Lines       10937    10937           
=======================================
+ Hits        10009    10010    +1     
+ Misses        928      927    -1     

Impacted Files	Coverage Δ
esmvalcore/preprocessor/_mask.py	85.89% <100.00%> (+0.42%)	⬆️

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

[axel-lauer]

I tested this bug fix and it works! Without the fix, values that are zero (0.0) are wrongly masked as missing values, see example of observed precipitation here (note e.g. parts of Antarctica):

With the fix, the zeros are kept as valid points, just as it should be:

Here is the test recipe used: recipe_masking_test.yml.txt

Thanks @schlunma for fixing this long-standing bug!

[valeriupredoi]

what an ancient bug! Cheers for spotting this @schlunma and cheers for fixing it 🍺

[schlunma]

@ESMValGroup/esmvaltool-coreteam anyone time to merge this? Thanks!

[bouweandela]

@schlunma Thanks for fixing this! A severe bug deserves an issue, could you please open it? That will make it much easier to find this problem and that it has been solved for anyone who encounters it. I will merge this now, so you can immediately close it again with a reference to this pull request.

[valeriupredoi]

@schlunma Thanks for fixing this! A severe bug deserves an issue, could you please open it? That will make it much easier to find this problem and that it has been solved for anyone who encounters it. I will merge this now, so you can immediately close it again with a reference to this pull request.

Good point, Bouwe! I wouldn't call it a severe bug myself, but heyho. More of an old, annoying horsefly-like bug 😁