Hi,
So, I read some time ago on the forum that research project might reject observations having accuracy as “not recorded”.
That bothers me a lot because I use a precise GPS(HDOP < 4m) and make sure that the location is recorded as precisely as possible for all my obs.
I almost exclusively use the website for my publications, and when I publish, all of them have accuracy “not recorded”.
I have over 10K obs and I don’t see myself updating them manually…
So… what are my options ?
Any suggestions welcome !
Thierry
1 Like
I’m not sure what the problem is here. I’ve checked a few of your observation, and they give accuracies of 3 m. So it sounds like that information is being stripped out when you download the observations? In that case I’d check the download page.
1 Like
You can select and edit observations in bulk (at least 200 or so at a time), using for example:
https://www.inaturalist.org/observations/thierrya?acc_below_or_unknown=4
(here it selects all your observations with their accuracy unset or lower than 4 meters) - use ‘batch edit’, ‘select all’, ‘edit selected’, then edit the accuracy value, ‘apply’ to all, then save the changes.
Just replace “acc_below_or_unknown=4” with “acc_above=4” to get to observations that have accuracy > 4 meters. And you can combine several arguments, e.g. “acc_above=4&acc_below=8”.
Note that ‘accuracy’ is loosely defined — and used — on iNaturalist (hdop/vdop? precision? ground-truth accuracy? assessed how?).
This bit of information is probably inconsistent between observers, and hard to interpret by interested parties. (Personally, I would have doubts about observations by non-geodesists claiming to always achieve a 3 m “accuracy”, unless the intended meaning and technical details are provided.)
Still, it allows drawing a pretty range radius on the maps, depicting ‘where’ the creature was (purportedly) observed. (Except if the observation or whole taxon is obscured, obviously.)
2 Likes
How can I find only observations without any accuracy set.
Sorry, but no: what you saw was me using the batch edit.
It feels a bit like cheating because I used plausible and honest values, but still not “real accuracy”
I used the bulk update, but not “at least 200 at a time”
It bugs out over batches of more than 100 records…
Still 100 at a time, it did the job.
Thanks also for the pointers to the url parameters, but they do not seem to work: [acc_below_or_unknown=4] returns observations where I manually did set the accuracy to 3m…
I might not be a geodesist but I’m a GPS freak and I’m pretty sure that when outdoors, I have an accuracy that is always under 5m: I always check the tracks on the computer to be sure of the pictures positioning.
Sorry but none at the moment: I spent the day updating all of them using the batch edit…
Then it’s a bug, the (website) upload function stripping the required GPS metadata bits from your files (e.g. the “GPSHPositioningError” EXIF tag, which is successfully interpreted as ‘accuracy’ by iNat, at the moment).
You should file a bug report in the appropriate section of this forum, if none exists already…
(I’m also a GPS freak btw ;))
That’s expected. 3<4.
Thanks, I’m going to investigate this.
1 Like
I found at least this old Bug Report: https://forum.inaturalist.org/t/gps-accuracy-not-captured-when-uploading-photos-taken-with-a-camera-with-a-built-in-gps/18861/23
But I think it should have been solved by now!? I regularly create observations on iNaturalist by uploading photos through the website, and both GPS coordinates and accuracy are correctly extracted from photo metadata (their tags, as reported by ‘ExifTool’, are GPSLatitude GPSLongitude and GPSHPositioningError).
This feels a bit counter intuitive to me: If I’m not mistaken (and I’m known to be, quite often…grinning:), we are talking of the GPSHPositioningError, where lower is better… so 3m is BETTER than 4m, no ?
Is there an URL parameter that would return only observation where accuracy is not set ?
Scratch this last question: I found it
acc=false
1 Like
To find observations where the accuracy value is below (= under, lower than) 4 meters:
?acc_below=4
In plain language one would say it is “more accurate” or with a “better/greater accuracy”, but the numerical value really is… smaller, lower. “1 is below 4”
It appears to work only with the “Observations” page, but not with the “Edit Observations” page.
However https://www.inaturalist.org/observations/thierrya?acc_below_or_unknown=1 does the job :)
1 Like
Many thanks ! This is EXACTLY what I needed !
This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.