[FEEDBACK] Moods breakdown question

[rombat]

I'm using the latest 3.1.0 version, with default settings.
I've just finished scanning my entire library (~115k unique songs), done a cleanup, etc.
I find odd that for a reasonably large and rather eclectic library, the moods breakdown ends up with shares being almost the same for all moods (less than 1% difference)

is it normal?

[NeptuneHub]

Thanks for sharing this. I investigate more (you wasn't the only talking about this) and was a wrong way to calculate the data for that graph that bring to unuseful number.

With this PR (wait around 60 minute for a new build):

• Multiple fix #828

Now each song is countend only ONE TIME for the mood that more rappresented the song (so the mood with the higher score). This give more miningful result.

Give a check and let me know.

[rombat]

Thanks, I'll try this and let you know

[rombat]

I don't know how accurate it is (I didn't think I had this much relaxed tracks, but that's fine), but at least it's more coherent :) Thanks !

[NeptuneHub]

I think that now work! thanks for feedback, going to close this issue.

[github-actions]

We've closed the issue now, thanks for having starred AudioMuse-AI and supporting self-hosted music here on GitHub! It really helps the project grow!