false positives issues find with ```analyser_merge_power_tower_FR.py``` reappear

[phd0]

Good morning,
I have a doubt about taking into account false positives on RTE poles and pylons in France.
If on "Report the issue as improper." it seems to reappear 15 days later, after re-executing the python script /analyses/analyse_merge_power_tower_FR.py
Indeed, the last execution dates from 08/08/2023, and I have the impression of finding the elements to be retouched in Osmose.

and looking here, I see that my changes from 15 days ago reappear then what have been corrected on ID and validated on Osmose.

In addition, the 10m positioning tolerance of the pylon seems too strict to me at first. This does not make it possible to differentiate the missing pylons from the thawed pylons shifted (or rebuilt a few meters further).
I think that if we really want to reduce the number of issules, we have to work in stages, and start by adding all the missing pylons. Then we can take care of the pylons not quite in place.

[frodrigo]

All the checks are made every days. Closing the Osmose issue just remove it until the next update. Rightly fixed osmose issues should not reaper as expected.

The conflation distance is fixed to 10m. https://github.com/osm-fr/osmose-backend/blob/master/analysers/analyser_merge_power_tower_FR.py#L51

If in some case the positions is far away, you can mark the osmose issue as false positive.

[flacombe]

Hello

First of all, 10m conflation distance should be preserved due to these situations:
https://twitter.com/RTE_Est/status/1684545241683050496?s=20

Some towers were rebuilt very next to the original position so we need to detect such changes with a comparable conflation distance.

Secondly, reported false positive come back for no known reason.
Let's make a test:
On this node https://www.openstreetmap.org/node/1372343866
The current issue https://osmose.openstreetmap.fr/fr/issue/df5dad45-3e39-f56e-1762-455e17bf355e has just been flagged false positive.

GET https://osmose.openstreetmap.fr/api/0.3/issue/df5dad45-3e39-f56e-1762-455e17bf355e/false has got HTTP 200 with null response.

Then wait and see if it comes back with a different hash in a few days.

[flacombe]

Current graph comes up a few days ago with a little rise.
Are we sure this issue is completed @phd0 ?

https://osmose.openstreetmap.fr/fr/issues/graph.png?item=8290&start_date=2023-08

[phd0]

honestly, I relied solely on the examples above, which serve as witnesses to confirm or not, my question. They have not moved for several weeks, I conclude that it is OK.
The curve can be linked to new additions from the openData file: but how to be sure?

[flacombe]

It's true example above didn't show up yet.

Opendata file is still from decembre 23, 2022. No update so far
https://odre.opendatasoft.com/explore/dataset/pylones-rte/information/

I bet it's others false positives, particularly in south west which was completely blank a few days ago.

[phd0]

So, do you recommend re-opening the issue?

[flacombe]

Yes, keep it open please

[flacombe]

Targeted issue in the upside test on https://www.openstreetmap.org/node/1372343866 node didn't show up again so far.

I give it a supplementary try on two nodes:

• https://www.openstreetmap.org/node/1675479294 (https://osmose.openstreetmap.fr/fr/issue/48804168-c1a8-5c0c-3b36-0fbb4c6299a4)
• No osm node (https://osmose.openstreetmap.fr/fr/issue/258d16e0-dd3a-9e41-1164-5a8102138f8e)

These two are flagged as false positive started from today.