ArctosDB / arctos

Arctos is a museum collections management system
https://arctos.database.museum
60 stars 13 forks source link

Request - Make Arctos stats (system stats) accessible in an exportable table #8097

Open mkoo opened 1 week ago

mkoo commented 1 week ago

Describe what you're trying to do The immediate need is for the UCDavis Data science students to create Tableau visualizations for Arctos data (more on that later!). One set of their proposed data dashboard would be to graph the stats in https://arctos.database.museum/home.cfm and https://arctos.database.museum/info/sysstats.cfm

The stats could be in a dedicated stats api or other web accessible table or file, refreshed by whatever processes already are creating the tallies on the above pages. Just let them know but it seems pointless for the students to recreate the summary statistics from flat. If we can replace or compliment our stats pages with a better data dashboard all the better!

mkoo commented 1 week ago

Reducing urgency since the students can access CSV of the system stats for now.

dustymc commented 1 week ago

@mkoo these links have been available and I think you've found them:

https://arctos.database.museum/Admin/CSVAnyTable.cfm?tableName=cache_sysstats_global

https://arctos.database.museum/Admin/CSVAnyTable.cfm?tableName=cache_sysstats_coln

Those are already maintained by the scheduler.

I wrapped the public (I hope, please review) bits of collection up into an object, which you can pull from

https://arctos.database.museum/Admin/CSVAnyTable.cfm?tableName=cache_collection

That could be added to the scheduler for maintenance and made available wherever. (https://github.com/ArctosDB/arctos/issues/8096 will likely change what's possible.)

TODO:

Do we need to talk about "how public" this is? Eg if we load it to corral then it can all get picked up by whomever for whatever purpose (which seems fine to me, but ???)

API vs download - I think the latter makes sense, this is a whole package (unlike eg filtered_flat where they API can find things in filtered_flat by things which are not in filtered_flat - taxonomy that a collection doesn't prefer, intersections with spatial data, etc, etc, etc.).

Schedule, if we get there (monthly, probably).

?? anything else ?