corpusmusic / bb-cluster

Markov analysis of harmonic progressions and cluster analysis of song-level harmonic profiles in the McGill Billboard dataset.
https://medium.com/modeling-music
GNU General Public License v3.0
9 stars 1 forks source link

Write script that will generate TransProb tables by cluster #15

Closed kshaffer closed 8 years ago

kshaffer commented 8 years ago

Take the cluster lists output from kmean_cluster.py and use them to create new subtables from the songbysongtransprob.csv data. For each cardinality, create a new csv file for each cluster that contains the data from songbysongtransprob.csv for only those songs.

Result will be a folder of files:

cluster1_1.csv (all the songs) cluster2_1.csv (two cluster solution, song data from cluster 1) cluster2_2.csv (two cluster solution, song data from cluster 2) cluster3_1.csv (three cluster solution, song data from cluster 1) ... cluster15_15.csv (15 cluster solution, song data from cluster 15)