Modify bb_transprob_cluster_analysis.R to automatically run on all cluster data tables
kshaffer opened this issue · 0 comments
bb_transprob_cluster_analysis.R currently has a new manual section of code for each clustering solution in the 2016 class. Rewrite the code so that it loops over every csv file created by the cluster table generator (cluster1_1.csv, cluster2_1.csv, cluster2_2.csv, ... cluster15_15.csv) and then outputs a CSV file and a PNG image for each.
The output CSV file will contain a 12x12 table of average transitional probabilities for all songs in the cluster.
The output PNG file contains a plotted visualization of this data. (Though there are weighting issues with the current plot solution. Is there a better one? Do we need a plot? Maybe instead, we can manually create a harmonic syntax "map" for some of the more interesting clusters?)