summaryrefslogtreecommitdiff
path: root/wk7/pset/songs/answers.txt
blob: 58c5634fcd038ae68008d620943066c71210eda4 (plain)
1
2
3
4
5
6
7
I would take the average value of the energy, valence, and danceability of each
song, then calculate the moods based on those values

One downside of this is that it may cancel out extreme values, leading to values
just in the  middle, and an inaccurate mood given. A better way to do this would
be to find clumps of the most common values of each and use the two biggest clumps
to base the mood on.