Change Identify Modal mini-map display for large accuracies

do you mean a filter so obs with a low accuracy don’t appear with the obs to be identified? Perhaps if the accuracy radius is too large they should be automatically marked casual? I notice in this discussion that if the radius is too large an obs won’t appear on the species map: https://forum.inaturalist.org/t/observation-not-showing-up-on-map/8383

2 Likes

I think there should be a worning appearing when uploading if the radius fits the whole/almost whole world (I guess you can remember the continent at least).

2 Likes

there was a big debate about that on the Google Groups. For a while that was the case - observations with very high uncertainty were marked casual. Then some people didn’t want it that way and it was put back. Originally some filters were going to be added to help, but I’m not sure it ever happened. Then as maybe a sort of compromise they were excluded from the range maps when they were really really imprecise.

I was one of those who thought they should be marked casual, and it’s one of my iNat pet peeves. I guess in some cases the huge circles are useful for wide ranging animals, etc… but for plants it is a whole other story.

2 Likes

Having a filter is completely independent of how to handle the display of the map. Maybe that conversation should be in a separate thread.

2 Likes

These filters do exist:

acc= [true or false]
acc_above= [metre value]
acc_below= [metre value]

I’ve gotten confused by this before, I think it would be an improvement.

1 Like

Added an issue here: https://github.com/inaturalist/inaturalist/issues/2500

1 Like

Might it be useful to display in some way observations without an accuracy value? One only discovers that they dont have an value when one really zooms in to fine scale. A little “x” or something

An option that might not go down too well is to rather than colour in the entire map when one enters the accuracy resolution scale, is to remove the record from the map display: after all strictly we no longer know if they actually occur in the mapped area.

1 Like

Since that isn’t related to observations with a large accuracy radius, you’ll need to open a new feature request or general discussion.

1 Like

I noticed @pleary put in a fix for this a couple days ago, but it’s not behaving exactly like I expect it to. The mini-maps on the identify modal are for the most part exactly as they used to be – at zoom level 5 – but a few are super zoomed in and I can’t figure out why those ones and not others. For example, when viewed in the identify modal, this one and this one are as zoomed in as on their observation pages, to level 14 and 19 respectively.

image

On the other hand, while the default zoom on this observation page shows the whole very large accuracy circle, in the identify modal, it’s the same problem as my original post – the default level 5 zoom doesn’t show any of the accuracy circle.

image

image

My preferred behavior for the mini maps on the identify modal is to have the more zoomed out of either zoom level 5 or a zoom level that encompasses the whole accuracy circle. Zoom level 14 or 19 is, in my opinion, too close to be useful. Bonus points if the zoom encompasses the full accuracy value even if the location has been obscured to a smaller rectangle.

I think the issue with these is that if you use the old uploader or if you add the location manually from the edit observation page, whatever zoom level used in those instances becomes the standard zoom level for the observation, which is also used in Identify. So I think that’s expected behavior.

1 Like

Given the snooping I did on github, I would believe that is the expected behavior for the new code. But the new code didn’t solve the issue I started this thread with, and now that I’ve seen it in action, I would even argue it’s added a new problem – zooms so close they’re no longer useful.
I would personally like to see a zoom no closer than 5, but wider than 5 if necessary to encompass the full accuracy. I realize other people may want closer zooms, but I have a hard time imagining what identifier might find the second screenshot I posted above (from ponch1977) useful.

image
image
image
image

I think it’s less useful for people who are mostly identifying mobile animals, but when IDing plants, I usually have maps shown as satellite view in order to better see the habitat type and surroundings. In this case a lot more information is gleaned by the close zoom on satellite view. I can see the observation is in a sunny, recently disturbed area as opposed to an adjacent woodlot.

I do a lot of click click click click click click click click click 10 or more times to zoom in already and really appreciate it when users set the zoom closer intentionally. See feature request to transfer that existing functionality to the new uploader: https://forum.inaturalist.org/t/maintain-map-zoom-level-when-creating-observations/8195 and other request to set your own default zoom level preference: https://forum.inaturalist.org/t/set-the-default-google-maps-zoom-level-for-observations/6827

image
Related? Obscured obs should be more zoomed out, regardless of the observer’s selected zoom level, right?

So anyway, I don’t have a strong opinion on how the “large accuracy” observations that are the subject of this topic should appear on Identify (i.e. retain the zoom level but fill the screen the accuracy circle color vs. automatically zoom out), but I’d lean toward zooming out. For normal/small accuracy radius and geoprivacy=open observations, I do think observations should retain their manually set zoom level as chosen by the observer/user.

3 Likes

wait, you can set a default zoom? How did I not know that? I zoom in a lot too, for plants.

2 Likes

You already mentioned that above lol. This is a topic about the map display of these things. I can’t remember if there is a feature request to add that to the user interface or make it sticky, but for now just add acc_below= and a number in meters as @reuvenm said above.

1 Like

sorry. really serious sleep deprivation here… teething is happening.

2 Likes

To clarify – the context of this particular thread is on the usefulness of the mini-map to determine the location of the observation. I will revise my statement to: I have a hard time imagining what identifier might find the second screenshot I posted above (from ponch1977) useful to determine the location of the observation.
It may certainly be useful for determining properties of the location, e.g. whether it’s forested, whether it’s recently disturbed, etc., but unless you have some outside knowledge (e.g. you know where the observer posts from, you’ve filtered results to a small area, etc.), this observation could be in Maine or Arizona or Alaska.

image

1 Like

Ah, yes, I usually filter by state or county, so requiring learning information like that from the map is generally extraneous. I would think adding the Standard Place levels 1 and maybe 2 next to the level 0 (United States in this case) would fulfill that need.

@charlie good luck with your new teeth

4 Likes

hahaha well they aren’t mine, wish they were, would be easier that way… haha

1 Like

my cat is teething now. he may not be as loud as a human baby but he does spend a lot of time chewing on my hands. and face (ow)

1 Like