Sort by within a project is not sorting based on the totality of observations

Platform:

App version number, if a mobile app issue: Inaturalist Website

Browser, if a website issue:

URLs (aka web addresses) of any relevant observations or pages:

https://www.inaturalist.org/observations?project_id=89210

Screenshots of what you are seeing:

Description of problem:

Step 1: Within the project “Lepidoptera of the world” which I believe is an umbrella project, Select Observers and SortBySpecies.

Step 2:

Go to the latest shown positions (around 500th place)

You can notice that the 500th place with just 142 species seen is being put.

Step 3:

As an example. Look for my user “jorge_natura” I am not appearing within that list.

But if this time I try to filter all MY observations within the project.

I can see that I have 274 species count. But not appearing on the SortBy species list.

It seems to me that iNaturalist is cutting the table based on total observations at some quantity and using that partial fraction of data to sort by, which ends up cutting some of the real results within the project.

Ps. I just wanted to point out that this is somewhat interesting to notice because It may be cutting some interesting statistics, and cutting data by the quantity of observations instead of other metric, like the number of species seems rather arbitrary.

Yeah, even I have noticed this. It probably takes the first 500 observers sorted by their number of observations and then sorts that list by number of species, potentially excluding from the species leaderboard a observer with an observations rank more than 500 but a species rank less than 500. It’s just an interesting behavior, maybe sorting by observations is computationally easier and they don’t want tons of observers to be sorted by species because it probably takes a full look through all their observations while sorting by observation is just counting them. I’m fine with this behavior though.

Yeah I am sure they trying not having to sort a table of data with a million rows each time a user wants to. Perhaps they could try to use a somethat bigger threshold to cut the table IMO.

this seems to be it (and i think more specifically, it’s probably more expensive to sort by species count than by observation count), but i don’t think staff have ever confirmed that this is the intent. more information here: https://forum.inaturalist.org/t/cutoff-for-2023-top-500-exceeds-6-000-observations/47901/27.

correction: i think staff have said this is intentional: https://forum.inaturalist.org/t/incorrect-calculation-for-most-species-on-the-project-page/62501/4.

1 Like

This topic was automatically closed after 22 hours. New replies are no longer allowed.