CNC 2024 First and Notable analysis

General
1

As in the previous years, I’ve created a small analysis of the CNC Wien project, to highlight notable observations at city, country, continent and global levels, as well as making it easy to see which taxa were uniquely observed by each individual.

https://waterpigs.co.uk/inat/cnc-wien-2024.html

The code (unfortunately without much documentation so far) is available here https://github.com/barnabywalters/inaturalist-project-stats if you want to try running it yourself. I’m also happy to run the analysis for you, just let me know :)

5 Likes
2

Is there a way to get the export file for a project when you are not the admin of the project (or not even a member of the project)?

Edit: nevermind i figured it out. Either use the address
https://www.inaturalist.org/observations/export?projects%5B%5D= + the project name or in the export page itself you type the name of the project

3

Yep, that’s correct. You’ll need to also include the non-default column taxon_species_name in order for the script to work properly.

4

Could you offer us some hints about how to run the script (since there’s no documentation)?

5

I managed to run it by

  1. downloading the repository
  2. extracting to a folder
  3. Open code in a terminal (i was using visual code studio and had to install some modules i was missing)
  4. Export the .csv from your project at https://www.inaturalist.org/observations/export like we discussed above
  5. At terminal run the command py uniquely_observed_species.py your_CSV.csv > results.html (there’s some clues on how to use it with different options in the file uniquely_observed_species_noconfig.py)
  6. wait for output.

In my case that command didn’t work with an error saying

No such file or directory: 'data [csv name] config.yaml'

but it worked if i used the noconfig.py file instead

Although i got another error related to species with name containing a “-”

1 Like
6

Just a heads-up, this category will be hidden from public view in a few days. Would you like me to move this topic to the General category?

7

glad you managed to get it to work, but the noconfig version is very old and I’m surprised it works at all any more! Seeing as people are interested in running it themselves, I will add some actual documentation and a sample config for the latest version. I’ll also take out the parts which are hardcoded for Vienna :)

oh yes that would be great, thanks! This category seemed the most appropriate place, I didn’t realise it would disappear imminently

2 Likes
8

No worries! I’ve moved it.

9

Thanks for posting this, this is really cool!

Do you have a sample config.yaml? The output from uniquely_observed_species.py seems like it’d be nicer but I’m not sure what all I need in the config.yaml file. I’m also seeing a handful of bugs in the _noconfig.py version which it looks like you’ve fixed some of those in the with-config version:

There seems to be a bug where it tries to create urls with &page=1&page=2 which causes a 500 error in the iNaturalist API. Changing lines 43 and 44 to

paged_api_url = f"{api_url}&page={page}&ttl={ttl}"
jresp = requests.get(paged_api_url).json()

so the original api_url isn’t getting overwritten and appended to every time fixes the issue for me.

The notable species list is outputting lots of nans for taxon IDs and names. Can’t figure out why.

Also, the taxa links in the final output html all 404 for me, because they’re mostly appended with “.0” after the taxon ID (e.g. https://www.inaturalist.org/taxa/213892.0). It looks like you fixed this in the with-config version.

EDIT: Missed your last post saying _noconfig is super old and that you’re gonna add a sample config, my bad!

10

Your post is well timed, I spent the last hour working on the script and documentation to make it a bit easier to use. The older _api and _noconfig versions are gone, replaced with the version which actually works, now with documentation and an example config!

For anyone curious, here’s a report for the Kerala umbrella project: https://waterpigs.co.uk/inat/cnc-kerala-2024.html and an even newer one for Houston-Galveston: https://waterpigs.co.uk/inat/cnc-houston-2024.html

1 Like
11

This is a very useful tool to locate “notable-but-actually-not-that-much-because-this-is-not-from-this-continent” observations to clean up these projects. I’ve been using it on brazilian CNC projects with nice results.

I had to change the code a bit in order to get less unicode character errors by adding
, encoding='utf-8') to all “open” commands . These happened mostly with common names that had the “ā” letter (which is not even used in portuguese)

So thanks a lot!

3 Likes
12

ha ha yes, it unintentionally also works as a tool for quickly finding and correcting completely wrong IDs! Maybe I should change it to “Firsts, Notable, and Wildly Misidentifed Observations” :thinking:

interesting, what platform are you on? I’ve never had any issues with this on mac os or linux with e.g. German exports, but I can add this to the script if it’s helpful for others

1 Like
13

This is a very useful tool to locate “notable-but-actually-not-that-much-because-this-is-not-from-this-continent” observations to clean up these projects. I’ve been using it on brazilian CNC projects with nice results.

Yes! I also used the analysis to find several incorrect IDs for our project. It was very useful for that!

14

This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.