Ouranosinc / xclim / 23750608729

92%

push

github

More SPI/SPEI pre-approved dists & Fix `rv_continuous` bug (#2327)

<!--Please ensure the PR fulfills the following requirements! -->
<!-- If this is your first PR, make sure to add your details to the
AUTHORS.rst! -->
### Pull Request Checklist:
- [x] This PR addresses an already opened issue (for bug fixes /
features)
    - This PR fixes #2326 
- [x] Tests for the changes have been added (for bug fixes / features)
- [x] (If applicable) Documentation has been added / updated (for bug
fixes / features)
- [x] CHANGELOG.rst has been updated (with summary of main changes)
- [x] Link to issue (:issue:`number`) and pull request (:pull:`number`)
has been added

### What kind of change does this PR introduce?

* Fix bug preventing the use of `rv_continuous` distributions in
SPI/SPEI
* Add direct support for genextreme/lognorm in SPI/SPEI
* Fix a bug in `lognorm`'s `_fit_start` (like other cases, `loc0` should
not exceed `min(x)`)
* Don't allow the use of `genextreme` with method='APP'.

### Does this PR introduce a breaking change?
I'd like to remove the support for `method="APP"` with `genextreme`. I
can add a message to say why it was removed. I'd be enclined to simply
not allow previous procedure. Contrary to other `_fit_start` cases, the
shape param of `genextreme` doesn't use data for a first gross
estimation. I don't think this should be promoted as an approximately
correct fit. Other cases like `gamma` can still be defended on the other
hand (and in case of `gamma`, reproduce monocongo's approach)

### Other information:

20 of 22 new or added lines in 4 files covered. (90.91%)

7927 of 8628 relevant lines covered (91.88%)

6.27 hits per line