idx_id not set when adding agents. about mesa-geo

[Corvince]

Thanks for reporting, Not sure why I used idx_id at all, but the Rtree should be able to use the unique_id of the agents. I'll do the necessary changes

[bpmweel]

I'm not sure this solution will work in every case, because the unique_id of an agent can be anything, while rtree expects a long integer.

In my case my agents have a string as a unique id, which is loaded from a geojson file.

[Corvince]

Ah, that was probably the reason in the first place! Thanks for letting me know. I actually should have tested my own solution, but my local set-up broke and I haven't found the time to fix it.

If you have a suggestion please feel free to tell me or submit a PR. Setting idx_id as an agent attribute as originally implemented would of course work, but am not sure if is such a good idea (and if probably only as a private attribute _idx_id

[Corvince]

Ah, probably id(agent) should be sufficient as an index integer, no?!

[bpmweel]

Yes I agree.

I took a look at id(agent) and the documentation says This is an integer (or long integer) which is guaranteed to be unique and constant for this object during its lifetime which fits quite well with the use in mesa-geo.

I also like it more than setting an _idx_id as a side-effect.

[Corvince]

Ok I tried using id(agent). Could you please test it? I am still not able to run my own project atm 😜

[bpmweel]

It took me a while to get back to this, but I tested it today and it seems to work well for me. Thanks for the changes!

[Corvince]

Hah! I just remembered this issue today and wasn't sure if I ever fixed it. But glad I did and it worked ;)

Thanks for coming back to this.