Matching varieties

[Former user]

Dear Thomas,
This is a suggestion for a possible addition in future versions, regarding the matching of names with varieties (var.). At the moment, a name including var. are matched as a whole unit. For our uses of Taxonome, it would be very useful if the matching is done in two stages: first, try to match the whole name (including var. name), and if no match was found, match the basic name only (that is, the genus and species with the basic authority). For example, when the name 'Iris verna L. var. smalliana Fern. ex M.E. Edwards' can't be matched, then the name ' 'Iris verna L.' should be attempted.
In our experience, this makes much sense, since varieties are usually not a taxonomicaly - independent unit.
I'd like to hear your opinion about this, and if possible, to add this option to the program.
Thanks,
Lior

[Thomas Kluyver]

Hi Lior,

I agree, that does make sense. In fact, the option already exists, but by default it will only move up to the species for nominal subspecies or varieties (e.g. Iris verna var. verna). If you want it to do this for all subspecies and varieties, pass upgrade_subsp='all' to the matching function (docs). In the GUI, the option is called "Match any subspecies to species".

Let me know if that works for you!

[Thomas Kluyver]

P.S. I spent a bit of time working on Taxonome again recently. I'm updating all the web service wrappers to use the new collections API, and then hopefully I should be able to make a new release - with the SQLite taxon store - quite soon.

[Former user]

Hi, and thanks for the quick reply.
That sounds like a good solution. However, we were advised to apply this method for varieties only, and not for subspecies. Therefore, we consider making a local change to the code in order to customize it for our needs. Could you help us by directing to the region of code where you think the change should be done. Also, do you think that such a change will have other (unwanted) effects on the program?
Regarding the new version, it would be very useful for us if we could you SQLite tables instead of CSVs, so this is a very welcome upgrade.
Thanks again,
Lior

[Thomas Kluyver]

You'll need to look at the code starting from this line, in TaxaResource.select. You'll want to check name.rank == 'variety'. Or possibly != 'subspecies' (which would also allow it for f. (forma) names). I don't think that change would have any unexpected effects.

I wonder if there's a good way to let the user configure which subspecific ranks are upgraded. It gets a bit fiddly if we want to preserve the option to upgrade only nominal subspecies.

The SQLite store won't have the same flexibility as the CSV reader - it will expect to store and retrieve taxa from a particular table format. I hope that will work for your use cases.