vergauwenthomas / MetObs_toolkit

A toolkit for using non-traditional meteorological observations
https://vergauwenthomas.github.io/MetObs_toolkit/
MIT License
12 stars 4 forks source link

[JOSS] docs per version hosted by read the docs #451

Closed vergauwenthomas closed 6 months ago

vergauwenthomas commented 6 months ago

See comment by @ashwinvis:

@vergauwenthomas I am now fairly confident that the software works as you have described.

I have a few requests regarding the article, particularly the section Technical implementation.

Further reading and citations needed here Since MetObs-Toolkit implements quite a few methods, it would be important to link to other sources which may describe them in detail. If you think that these methods might be subject to change, it might be sufficient to cite the references in the online documentation and link this software paper to the documentation instead. See how scikit-image does it.

“55 to examine data types, irregular timestamps, max-min thresholds, repetitions criteria, spike tests, allowed variation in time windows and spatial tests” (Vergauwen et al., 2024, p. 3) (pdf)

These, IMHO, are quite involved methods. What are the formal basis behind these QC methods or are they based on intuition? Some reuses TITAN (titanlib) while the rest are implemented from the scratch (which means citing TITAN alone will not be sufficient). .

“59 ... gap filling is applied by using interpolation methods” (Vergauwen et al., 2024, p. 3) (pdf)

Gap filling and bias correction is another involved topic. What is the basis behind the methods used? Further reading and citations needed here.


When a reference is made to the documentation, there must be an online version-tracking of the documentation. In order to do that:

vergauwenthomas commented 6 months ago

Ref: https://github.com/openjournals/joss-reviews/issues/5916