nkululeko 0.82.3__tar.gz → 0.83.0__tar.gz

{nkululeko-0.82.3 → nkululeko-0.83.0}/CHANGELOG.md

@@ -1,6 +1,14 @@
 Changelog
 =========
 
+Version 0.83.0
+--------------
+* test module now prints out reports
+
+Version 0.82.4
+--------------
+* fixed bug in wavlm 
+
 Version 0.82.3
 --------------
 * fixed another audformat peculiarity to interprete time values as nanoseconds

{nkululeko-0.82.3/nkululeko.egg-info → nkululeko-0.83.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: nkululeko
-Version: 0.82.3
+Version: 0.83.0
 Summary: Machine learning audio prediction experiments based on templates
 Home-page: https://github.com/felixbur/nkululeko
 Author: Felix Burkhardt
@@ -333,6 +333,14 @@ F. Burkhardt, Johannes Wagner, Hagen Wierstorf, Florian Eyben and Björn Schulle
 Changelog
 =========
 
+Version 0.83.0
+--------------
+* test module now prints out reports
+
+Version 0.82.4
+--------------
+* fixed bug in wavlm 
+
 Version 0.82.3
 --------------
 * fixed another audformat peculiarity to interprete time values as nanoseconds

{nkululeko-0.82.3 → nkululeko-0.83.0}/nkululeko/constants.py

@@ -1,2 +1,2 @@
-VERSION="0.82.3"
+VERSION="0.83.0"
 SAMPLING_RATE = 16000

{nkululeko-0.82.3 → nkululeko-0.83.0}/nkululeko/experiment.py

@@ -23,7 +23,7 @@ from nkululeko.plots import Plots
 from nkululeko.reporting.report import Report
 from nkululeko.runmanager import Runmanager
 from nkululeko.scaler import Scaler
-from nkululeko.test_predictor import Test_predictor
+from nkululeko.test_predictor import TestPredictor
 from nkululeko.utils.util import Util
 
 
@@ -672,7 +672,7 @@ class Experiment:
     def predict_test_and_save(self, result_name):
         model = self.runmgr.get_best_model()
         model.set_testdata(self.df_test, self.feats_test)
-        test_predictor = Test_predictor(
+        test_predictor = TestPredictor(
             model, self.df_test, self.label_encoder, result_name
         )
         test_predictor.predict_and_store()

{nkululeko-0.82.3 → nkululeko-0.83.0}/nkululeko/feat_extract/feats_hubert.py

@@ -6,23 +6,26 @@
 
 import os
 
-import audeer
-import nkululeko.glob_conf as glob_conf
 import pandas as pd
 import torch
 import torchaudio
-from audformat.utils import map_file_path
-from nkululeko.feat_extract.featureset import Featureset
 from tqdm import tqdm
-from transformers import HubertModel, Wav2Vec2FeatureExtractor
+from transformers import HubertModel
+from transformers import Wav2Vec2FeatureExtractor
+
+from nkululeko.feat_extract.featureset import Featureset
+import nkululeko.glob_conf as glob_conf
 
 
 class Hubert(Featureset):
-    """Class to extract HuBERT embedding)"""
+    """Class to extract HuBERT embedding)."""
 
     def __init__(self, name, data_df, feat_type):
-        """Constructor. is_train is needed to distinguish from test/dev sets,
-        because they use the codebook from the training"""
+        """Constructor.
+
+        Is_train is needed to distinguish from test/dev sets,
+        because they use the codebook from the training.
+        """
         super().__init__(name, data_df, feat_type)
         # check if device is not set, use cuda if available
         cuda = "cuda" if torch.cuda.is_available() else "cpu"
@@ -61,16 +64,12 @@ class Hubert(Featureset):
         """Extract the features or load them from disk if present."""
         store = self.util.get_path("store")
         storage = f"{store}{self.name}.pkl"
-        extract = self.util.config_val(
-            "FEATS", "needs_feature_extraction", False
-        )
+        extract = self.util.config_val("FEATS", "needs_feature_extraction", False)
         no_reuse = eval(self.util.config_val("FEATS", "no_reuse", "False"))
         if extract or no_reuse or not os.path.isfile(storage):
             if not self.model_initialized:
                 self.init_model()
-            self.util.debug(
-                "extracting Hubert embeddings, this might take a while..."
-            )
+            self.util.debug("extracting Hubert embeddings, this might take a while...")
             emb_series = pd.Series(index=self.data_df.index, dtype=object)
             length = len(self.data_df.index)
             for idx, (file, start, end) in enumerate(
@@ -84,9 +83,7 @@ class Hubert(Featureset):
                 assert sampling_rate == 16000
                 emb = self.get_embeddings(signal, sampling_rate, file)
                 emb_series.iloc[idx] = emb
-            self.df = pd.DataFrame(
-                emb_series.values.tolist(), index=self.data_df.index
-            )
+            self.df = pd.DataFrame(emb_series.values.tolist(), index=self.data_df.index)
             self.df.to_pickle(storage)
             try:
                 glob_conf.config["DATA"]["needs_feature_extraction"] = "false"

{nkululeko-0.82.3 → nkululeko-0.83.0}/nkululeko/feat_extract/feats_wavlm.py

@@ -4,27 +4,32 @@
 
 import os
 
-import nkululeko.glob_conf as glob_conf
 import pandas as pd
 import torch
 import torchaudio
-from nkululeko.feat_extract.featureset import Featureset
 from tqdm import tqdm
-from transformers import Wav2Vec2FeatureExtractor, WavLMModel
+from transformers import Wav2Vec2FeatureExtractor
+from transformers import WavLMModel
+
+from nkululeko.feat_extract.featureset import Featureset
+import nkululeko.glob_conf as glob_conf
 
 
 class Wavlm(Featureset):
-    """Class to extract WavLM embedding)"""
+    """Class to extract WavLM embedding)."""
+
+    def __init__(self, name, data_df, feats_type):
+        """Constructor.
 
-    def __init__(self, name, data_df, feat_type):
-        """Constructor. is_train is needed to distinguish from test/dev sets,
-        because they use the codebook from the training"""
-        super().__init__(name, data_df)
+        Is_train is needed to distinguish from test/dev sets,
+        because they use the codebook from the training.
+        """
+        super().__init__(name, data_df, feats_type)
         # check if device is not set, use cuda if available
         cuda = "cuda" if torch.cuda.is_available() else "cpu"
         self.device = self.util.config_val("MODEL", "device", cuda)
         self.model_initialized = False
-        self.feat_type = feat_type
+        self.feat_type = feats_type
 
     def init_model(self):
         # load model
@@ -59,7 +64,9 @@ class Wavlm(Featureset):
                     frame_offset=int(start.total_seconds() * 16000),
                     num_frames=int((end - start).total_seconds() * 16000),
                 )
-                assert sampling_rate == 16000, f"sampling rate should be 16000 but is {sampling_rate}"
+                assert (
+                    sampling_rate == 16000
+                ), f"sampling rate should be 16000 but is {sampling_rate}"
                 emb = self.get_embeddings(signal, sampling_rate, file)
                 emb_series.iloc[idx] = emb
             self.df = pd.DataFrame(emb_series.values.tolist(), index=self.data_df.index)

{nkululeko-0.82.3 → nkululeko-0.83.0}/nkululeko/modelrunner.py

@@ -2,18 +2,16 @@
 
 import pandas as pd
 
-from nkululeko.utils.util import Util
 from nkululeko import glob_conf
-import nkululeko.glob_conf as glob_conf
+from nkululeko.utils.util import Util
 
 
 class Modelrunner:
-    """
-    Class to model one run
-    """
+    """Class to model one run."""
 
     def __init__(self, df_train, df_test, feats_train, feats_test, run):
-        """Constructor setting up the dataframes
+        """Constructor setting up the dataframes.
+
         Args:
             df_train: train dataframe
             df_test: test dataframe

{nkululeko-0.82.3 → nkululeko-0.83.0}/nkululeko/reporter.py

@@ -3,7 +3,6 @@
 This module contains the Reporter class which is responsible for generating reports.
 """
 
-
 import ast
 import glob
 import json
@@ -12,16 +11,14 @@ import math
 import matplotlib.pyplot as plt
 import numpy as np
 from scipy.stats import pearsonr
-from sklearn.metrics import (
-    ConfusionMatrixDisplay,
-    accuracy_score,
-    classification_report,
-    confusion_matrix,
-    mean_absolute_error,
-    mean_squared_error,
-    r2_score,
-    recall_score,
-)
+from sklearn.metrics import ConfusionMatrixDisplay
+from sklearn.metrics import accuracy_score
+from sklearn.metrics import classification_report
+from sklearn.metrics import confusion_matrix
+from sklearn.metrics import mean_absolute_error
+from sklearn.metrics import mean_squared_error
+from sklearn.metrics import r2_score
+from sklearn.metrics import recall_score
 from sklearn.utils import resample
 
 import nkululeko.glob_conf as glob_conf
@@ -70,11 +67,9 @@ class Reporter:
             else:
                 # regression experiment
                 if self.measure == "mse":
-                    self.result.test = mean_squared_error(
-                        self.truths, self.preds)
+                    self.result.test = mean_squared_error(self.truths, self.preds)
                 elif self.measure == "mae":
-                    self.result.test = mean_absolute_error(
-                        self.truths, self.preds)
+                    self.result.test = mean_absolute_error(self.truths, self.preds)
                 elif self.measure == "ccc":
                     self.result.test = self.ccc(self.truths, self.preds)
                     if math.isnan(self.result.test):
@@ -115,7 +110,7 @@ def plot_per_speaker(self, result_df, plot_name, function):
     Args:
         * result_df: a pandas dataframe with columns: preds, truths and speaker
         * plot_name: a string with the name of the plot
-        * function: a string with the function to use for each speaker, 
+        * function: a string with the function to use for each speaker,
         can be 'mode' or 'mean'
 
     Returns:
@@ -185,8 +180,7 @@ def plot_per_speaker(self, result_df, plot_name, function):
             reg_res = f", {self.MEASURE}: {self.result.test:.3f}"
 
         if epoch != 0:
-            plt.title(
-                f"Confusion Matrix, UAR: {uar:.3f}{reg_res}, Epoch: {epoch}")
+            plt.title(f"Confusion Matrix, UAR: {uar:.3f}{reg_res}, Epoch: {epoch}")
         else:
             plt.title(f"Confusion Matrix, UAR: {uar:.3f}{reg_res}")
         img_path = f"{fig_dir}{plot_name}.{self.format}"
@@ -265,8 +259,7 @@ def plot_per_speaker(self, result_df, plot_name, function):
         import imageio
 
         fig_dir = self.util.get_path("fig_dir")
-        filenames = glob.glob(
-            fig_dir + f"{self.util.get_plot_name()}*_?_???_cnf.png")
+        filenames = glob.glob(fig_dir + f"{self.util.get_plot_name()}*_?_???_cnf.png")
         images = []
         for filename in filenames:
             images.append(imageio.imread(filename))
@@ -322,8 +315,7 @@ def plot_per_speaker(self, result_df, plot_name, function):
         var_pred = np.var(prediction, 0)
         v_pred = prediction - mean_pred
         v_gt = ground_truth - mean_gt
-        cor = sum(v_pred * v_gt) / \
-            (np.sqrt(sum(v_pred**2)) * np.sqrt(sum(v_gt**2)))
+        cor = sum(v_pred * v_gt) / (np.sqrt(sum(v_pred**2)) * np.sqrt(sum(v_gt**2)))
         sd_gt = np.std(ground_truth)
         sd_pred = np.std(prediction)
         numerator = 2 * cor * sd_gt * sd_pred

{nkululeko-0.82.3 → nkululeko-0.83.0}/nkululeko/reporting/reporter.py

@@ -55,6 +55,7 @@ class Reporter:
         self.run = run
         self.epoch = epoch
         self.__set_measure()
+        self.filenameadd = ""
         self.cont_to_cat = False
         if len(self.truths) > 0 and len(self.preds) > 0:
             if self.util.exp_is_classification():
@@ -206,7 +207,7 @@ class Reporter:
                 f"Confusion Matrix, UAR: {uar_str} "
                 + f"(+-{up_str}/{low_str}) {reg_res}"
             )
-        img_path = f"{fig_dir}{plot_name}.{self.format}"
+        img_path = f"{fig_dir}{plot_name}{self.filenameadd}.{self.format}"
         plt.savefig(img_path)
         fig.clear()
         plt.close(fig)
@@ -228,14 +229,17 @@ class Reporter:
         )
         # print(rpt)
         self.util.debug(rpt)
-        file_name = f"{res_dir}{self.util.get_exp_name()}_conf.txt"
+        file_name = f"{res_dir}{self.util.get_exp_name()}{self.filenameadd}_conf.txt"
         with open(file_name, "w") as text_file:
             text_file.write(rpt)
 
+    def set_filename_add(self, my_string):
+        self.filenameadd = f"_{my_string}"
+
     def print_results(self, epoch):
         """Print all evaluation values to text file."""
         res_dir = self.util.get_path("res_dir")
-        file_name = f"{res_dir}{self.util.get_exp_name()}_{epoch}.txt"
+        file_name = f"{res_dir}{self.util.get_exp_name()}_{epoch}{self.filenameadd}.txt"
         if self.util.exp_is_classification():
             labels = glob_conf.labels
             try:

{nkululeko-0.82.3 → nkululeko-0.83.0}/nkululeko/test_predictor.py

@@ -1,21 +1,25 @@
-""" test_predictor.py
+"""test_predictor.py.
+
     Predict targets from a model and save as csv file.
 
 """
 
-import nkululeko.glob_conf as glob_conf
-from nkululeko.utils.util import Util
+import ast
+
+import numpy as np
 import pandas as pd
+from sklearn.preprocessing import LabelEncoder
+
 from nkululeko.data.dataset import Dataset
 from nkululeko.feature_extractor import FeatureExtractor
+import nkululeko.glob_conf as glob_conf
 from nkululeko.scaler import Scaler
-import numpy as np
-from sklearn.preprocessing import LabelEncoder
+from nkululeko.utils.util import Util
 
 
-class Test_predictor:
+class TestPredictor:
     def __init__(self, model, orig_df, labenc, name):
-        """Constructor setting up name and configuration"""
+        """Constructor setting up name and configuration."""
         self.model = model
         self.orig_df = orig_df
         self.label_encoder = labenc
@@ -49,7 +53,14 @@ class Test_predictor:
             df[self.target] = labelenc.inverse_transform(predictions.tolist())
             df.to_csv(self.name)
         else:
+            test_dbs = ast.literal_eval(glob_conf.config["DATA"]["tests"])
+            test_dbs_string = "_".join(test_dbs)
             predictions = self.model.get_predictions()
+            report = self.model.predict()
+            report.set_filename_add(f"test-{test_dbs_string}")
+            self.util.print_best_results([report])
+            report.plot_confmatrix(self.util.get_plot_name(), 0)
+            report.print_results(0)
             # print(predictions)
             # df = pd.DataFrame(index=self.orig_df.index)
             # df["speaker"] = self.orig_df["speaker"]

{nkululeko-0.82.3 → nkululeko-0.83.0/nkululeko.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: nkululeko
-Version: 0.82.3
+Version: 0.83.0
 Summary: Machine learning audio prediction experiments based on templates
 Home-page: https://github.com/felixbur/nkululeko
 Author: Felix Burkhardt
@@ -333,6 +333,14 @@ F. Burkhardt, Johannes Wagner, Hagen Wierstorf, Florian Eyben and Björn Schulle
 Changelog
 =========
 
+Version 0.83.0
+--------------
+* test module now prints out reports
+
+Version 0.82.4
+--------------
+* fixed bug in wavlm 
+
 Version 0.82.3
 --------------
 * fixed another audformat peculiarity to interprete time values as nanoseconds