0.92.1

CHANGELOG.md

@@ -1,6 +1,10 @@
 Changelog
 =========
 
+Version 0.92.1
+--------------
+* added a speaker plot to pyannote results 
+
 Version 0.92.0
 --------------
 * added first version of automatic speaker prediction/segmentation

nkululeko/constants.py

@@ -1,2 +1,2 @@
-VERSION="0.92.0"
+VERSION="0.92.1"
 SAMPLING_RATE = 16000

nkululeko/plots.py

@@ -4,14 +4,14 @@
 import matplotlib.pyplot as plt
 import numpy as np
 import pandas as pd
-import seaborn as sns
 from scipy import stats
+import seaborn as sns
 from sklearn.manifold import TSNE
 
 import nkululeko.glob_conf as glob_conf
-import nkululeko.utils.stats as su
 from nkululeko.reporting.defines import Header
 from nkululeko.reporting.report_item import ReportItem
+import nkululeko.utils.stats as su
 from nkululeko.utils.util import Util
 
 
@@ -30,8 +30,6 @@ def plot_distributions_speaker(self, df):
             df_speaker["samplenum"] = df_speaker.shape[0]
             df_speakers = pd.concat([df_speakers, df_speaker.head(1)])
         # plot the distribution of samples per speaker
-        # one up because of the runs
-        fig_dir = self.util.get_path("fig_dir") + "../"
         self.util.debug("plotting samples per speaker")
         if "gender" in df_speakers:
             filename = "samples_value_counts"
@@ -319,6 +317,31 @@ def plot_durations(self, df, filename, sample_selection, caption=""):
         img_path = f"{fig_dir}{filename}_{sample_selection}.{self.format}"
         plt.savefig(img_path)
         plt.close(fig)
+        self.util.debug(f"plotted durations to {img_path}")
+        glob_conf.report.add_item(
+            ReportItem(
+                Header.HEADER_EXPLORE,
+                caption,
+                title,
+                img_path,
+            )
+        )
+
+    def plot_speakers(self, df, sample_selection):
+        filename = "speakers"
+        caption = "speakers"
+        # one up because of the runs
+        fig_dir = self.util.get_path("fig_dir") + "../"
+        sns.set_style("whitegrid")  # Set style for chart
+        ax = df["speaker"].value_counts().plot(kind="pie")
+        title = f"Speaker distr. for {sample_selection} {df.shape[0]}."
+        ax.set_title(title)
+        fig = ax.figure
+        # plt.tight_layout()
+        img_path = f"{fig_dir}{filename}_{sample_selection}.{self.format}"
+        plt.savefig(img_path)
+        plt.close(fig)
+        self.util.debug(f"plotted speakers to {img_path}")
         glob_conf.report.add_item(
             ReportItem(
                 Header.HEADER_EXPLORE,

nkululeko/segment.py

@@ -63,7 +63,7 @@ def main():
     # segment
     segmented_file = util.config_val("SEGMENT", "result", "segmented.csv")
 
-    segmenter = util.config_val("SEGMENT", "method", "silero")
+    method = util.config_val("SEGMENT", "method", "silero")
     sample_selection = util.config_val("SEGMENT", "sample_selection", "all")
     if sample_selection == "all":
         df = pd.concat([expr.df_train, expr.df_test])
@@ -76,19 +76,19 @@ def main():
             f"unknown segmentation selection specifier {sample_selection},"
             " should be [all | train | test]"
         )
-    util.debug(f"segmenting {sample_selection}: {df.shape[0]} samples with {segmenter}")
-    if segmenter == "silero":
+    util.debug(f"segmenting {sample_selection}: {df.shape[0]} samples with {method}")
+    if method == "silero":
         from nkululeko.segmenting.seg_silero import Silero_segmenter
 
         segmenter = Silero_segmenter()
         df_seg = segmenter.segment_dataframe(df)
-    elif segmenter == "pyannote":
+    elif method == "pyannote":
         from nkululeko.segmenting.seg_pyannote import Pyannote_segmenter
 
         segmenter = Pyannote_segmenter(config)
         df_seg = segmenter.segment_dataframe(df)
     else:
-        util.error(f"unknown segmenter: {segmenter}")
+        util.error(f"unknown segmenter: {method}")
 
     def calc_dur(x):
         starts = x[1]
@@ -110,6 +110,9 @@ def calc_dur(x):
     plots.plot_durations(
         df_seg, "segmented_durations", sample_selection, caption="Segmented durations"
     )
+    if method == "pyannote":
+        plots.plot_speakers(df_seg, sample_selection)
+
     print("")
     # remove encoded labels
     target = util.config_val("DATA", "target", None)