def main(df, args):
df = pandas.read_hdf('data/sonorities/sonorities-by-measure.hd5', "sonorities")
report_similarity_by_categories(df)
# report_similarities_by_year(df)
pcset_series = music.get_pcset_table()
pcset_diff_args = report_pcset_similarity_difference(pcset_series)
reports.make_json_file(pcset_diff_args, "categories-pcset-difference")
def main(df, args):
method = "interval_vector"
labels = ["Piece", "Sonorities"]
p = 0.01 # KS
eps = 0.05 # DBSCAN
h = df[df["composer"] == "haydn"]
m = df[df["composer"] == "mozart"]
b = df[df["composer"] == "beethoven"]
hmb = filter_classical_composers(df)
freq_df = get_frequency(hmb, method)
dist_df = frequency_matrix(freq_df, _distance)
ks_df = frequency_matrix(freq_df, _ks_test, {"p": 0.01})
# Amount
classical = different_sonorities_amount(hmb, method).to_javascript()
haydn = different_sonorities_amount(h, method).to_javascript()
mozart = different_sonorities_amount(m, method).to_javascript()
beethoven = different_sonorities_amount(b, method).to_javascript()
# Distance distribution
dist_mean = dist_df.mean()
dist_var = SpecialDataFrame(dist_df.var(), columns=["Distance Variance"])
normalized_dist_mean = SpecialDataFrame((dist_mean - dist_mean.mean()) / dist_mean.std(), columns=["Distance Mean"])
histogram_dist = SpecialDataFrame(dist_df.sum(), columns=["Euclidean Distance"])
ks_sum = SpecialDataFrame(ks_df.sum(), columns=["False KS Sum"])
amount_args = [
_chart(haydn, "Haydn quartets"),
_chart(mozart, "Mozart quartets"),
_chart(beethoven, "Beethoven quartets"),
_chart(classical, "Haydn, Mozart and Beethoven quartets"),
]
distance_heatmap_args = [
{
"data": dist_df.to_javascript(),
"chart-type": "BioHeatMap",
"options": {
"title": "Distance between quartets",
"mapWidth": 700,
"mapHeight": 500,
# bioheatmap options
"numberOfColors": 100,
"startColor": {"r": 255, "g": 255, "b": 255, "a": 1},
"endColor": {"r": 0, "g": 128, "b": 255, "a": 1},
"passThroughBlack": False,
"drawBorder": False,
},
},
{
"data": ks_df.to_javascript(True),
"chart-type": "BioHeatMap",
"options": {
"title": "Kolmogorov-Smirnov test between quartets",
"mapWidth": 700,
"mapHeight": 500,
# bioheatmap options
"numberOfColors": 100,
"startColor": {"r": 255, "g": 255, "b": 255, "a": 1},
"endColor": {"r": 0, "g": 128, "b": 255, "a": 1},
"passThroughBlack": "false",
"drawBorder": "false",
},
},
]
distance_distribution_args = [
{
"data": normalized_dist_mean.to_javascript(),
"chart-type": "ColumnChart",
"options": {
"title": "Normalized mean distance to quartets",
"hAxis": {"title": "Quartet name"},
"vAxis": {"title": "Euclidean distance (normalized by standard deviation)"},
},
},
{
"data": histogram_dist.to_javascript(),
"chart-type": "Histogram",
"options": {
"title": "Distance sum to quartets (histogram)",
"hAxis": {"title": "Euclidean distance (sum)"},
"vAxis": {"title": "Number of quartets"},
},
},
{
"data": dist_var.to_javascript(),
"chart-type": "ColumnChart",
"options": {
"title": "Distance variance to quartets",
"hAxis": {"title": "Quartet name"},
"vAxis": {"title": "Euclidean distance variance"},
},
},
{
"data": dist_var.to_javascript(),
"chart-type": "Histogram",
"options": {
"title": "Distance variance to quartets (histogram)",
"hAxis": {"title": "Euclidean distance (variance)"},
"vAxis": {"title": "Number of quartets"},
},
},
{
"data": ks_sum.to_javascript(),
"chart-type": "ColumnChart",
"options": {
"title": "False KS 2-sample sum",
"hAxis": {"title": "Quartet"},
"vAxis": {"title": "Number of False KS"},
},
},
{
"data": ks_sum.to_javascript(),
"chart-type": "Histogram",
"options": {
"title": "False KS 2-sample sum (histogram)",
"hAxis": {"title": "Number of False KS"},
"vAxis": {"title": "Number of quartets"},
},
},
]
distance_to_interval_vector_args = [
_chart_distance(sonorities_amount_and_distance(h).to_javascript(), "Haydn sonorities"),
_chart_distance(sonorities_amount_and_distance(m).to_javascript(), "Mozart sonorities"),
_chart_distance(sonorities_amount_and_distance(b).to_javascript(), "Beethoven sonorities"),
_chart_distance(sonorities_amount_and_distance(hmb).to_javascript(), "Haydn, Mozart and Beethoven sonorities"),
]
reports.make_json_file(amount_args, "sonorities-amount-timeline")
reports.make_json_file(distance_heatmap_args, "sonorities-distance-frequency-heatmap")
reports.make_json_file(distance_distribution_args, "sonorities-distance-frequency-distribution")
reports.make_json_file(distance_to_interval_vector_args, "sonorities-distance")
# cluster table
cluster_to_csv(freq_df.T, dist_df, hmb, eps)
