nkululeko 0.74.2__py3-none-any.whl → 0.74.6__py3-none-any.whl

nkululeko/augment.py

@@ -53,17 +53,21 @@ def main(src_dir):
     got_one = False
     if augmentings:
         augmentings = ast.literal_eval(augmentings)
-        def_ret = pd.DataFrame()
+        results = []
         if "traditional" in augmentings:
-            df_ret = expr.augment()
+            df1 = expr.augment()
+            results.append(df1)
             got_one = True
         if "random_splice" in augmentings:
-            df_ret = pd.concat([def_ret, expr.random_splice()])
+            df2 = expr.random_splice()
+            results.append(df2)
             got_one = True
     if not augmentings:
         util.error("no augmentation selected")
     if not got_one:
         util.error(f"invalid augmentation(s): {augmentings}")
+    df_ret = pd.DataFrame()
+    df_ret = pd.concat(results)
     # remove encoded labels
     target = util.config_val("DATA", "target", "emotion")
     if "class_label" in df_ret.columns:

nkululeko/augmenting/augmenter.py

@@ -30,10 +30,9 @@ class Augmenter:
         self.audioment = Compose(
             [
                 AddGaussianNoise(min_amplitude=0.001, max_amplitude=0.015, p=0.5),
-                # AddGaussianSNR(min_snr_db=10, max_snr_db=40, p=0.5),
                 TimeStretch(min_rate=0.8, max_rate=1.25, p=0.5),
                 PitchShift(min_semitones=-4, max_semitones=4, p=0.5),
-                Shift(min_fraction=-0.5, max_fraction=0.5, p=0.5),
+                Shift(p=0.5),
             ]
         )
 

nkululeko/constants.py

@@ -1,2 +1,2 @@
-VERSION="0.74.2"
+VERSION="0.74.6"
 SAMPLING_RATE = 16000

nkululeko/demo_predictor.py

@@ -42,6 +42,13 @@ class Demo_predictor:
 
     def predict_signal(self, signal, sr):
         features = self.feature_extractor.extract_sample(signal, sr)
+        scale_feats = self.util.config_val("FEATS", "scale", False)
+        if scale_feats:
+            from sklearn.preprocessing import StandardScaler
+
+            scaler = StandardScaler()
+            features = scaler.fit_transform(features)
+        features = np.nan_to_num(features)
         result_dict = self.model.predict_sample(features)
         keys = result_dict.keys()
         if self.label_encoder is not None:

nkululeko/experiment.py

@@ -100,12 +100,14 @@ class Experiment:
         dbs = ",".join(list(self.datasets.keys()))
         labels = self.util.config_val("DATA", "labels", False)
         if labels:
-            labels = ast.literal_eval(labels)
+            self.labels = ast.literal_eval(labels)
             self.util.debug(f"Target labels (from config): {labels}")
         else:
-            labels = list(next(iter(self.datasets.values())).df[self.target].unique())
+            self.labels = list(
+                next(iter(self.datasets.values())).df[self.target].unique()
+            )
             self.util.debug(f"Target labels (from database): {labels}")
-        glob_conf.set_labels(labels)
+        glob_conf.set_labels(self.labels)
         self.util.debug(f"loaded databases {dbs}")
 
     def _import_csv(self, storage):
@@ -589,6 +591,7 @@ class Experiment:
         if save:
             # save the experiment for future use
             self.save(self.util.get_save_name())
+            # self.save_onnx(self.util.get_save_name())
 
         # self.__collect_reports()
         self.util.print_best_results(self.reports)
@@ -667,6 +670,7 @@ class Experiment:
         tmp_dict = pickle.load(f)
         f.close()
         self.__dict__.update(tmp_dict)
+        glob_conf.set_labels(self.labels)
 
     def save(self, filename):
         try:
@@ -675,3 +679,17 @@ class Experiment:
             f.close()
         except (AttributeError, TypeError, RuntimeError) as error:
             self.util.warn(f"Save experiment: Can't pickle local object: {error}")
+
+    def save_onnx(self, filename):
+        # export the model to onnx
+        model = self.runmgr.get_best_model()
+        if model.is_ANN():
+            print("converting to onnx from torch")
+        else:
+            from skl2onnx import to_onnx
+
+            print("converting to onnx from sklearn")
+        # save the rest
+        f = open(filename, "wb")
+        pickle.dump(self.__dict__, f)
+        f.close()

nkululeko/feat_extract/feats_praat.py

@@ -2,6 +2,7 @@
 from nkululeko.feat_extract.featureset import Featureset
 import os
 import pandas as pd
+import numpy as np
 import nkululeko.glob_conf as glob_conf
 from nkululeko.feat_extract import feinberg_praat
 import ast
@@ -23,14 +24,10 @@ class Praatset(Featureset):
         store = self.util.get_path("store")
         store_format = self.util.config_val("FEATS", "store_format", "pkl")
         storage = f"{store}{self.name}.{store_format}"
-        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):
-            self.util.debug(
-                "extracting Praat features, this might take a while..."
-            )
+            self.util.debug("extracting Praat features, this might take a while...")
             self.df = feinberg_praat.compute_features(self.data_df.index)
             self.df = self.df.set_index(self.data_df.index)
             for i, col in enumerate(self.df.columns):
@@ -53,36 +50,25 @@ class Praatset(Featureset):
         self.df = self.df.astype(float)
 
     def extract_sample(self, signal, sr):
-        self.util.error(
-            "feats_praat: extracting single samples not implemented yet"
-        )
-        feats = None
-        return feats
+        import audiofile, audformat
 
-    # def filter(self):
-    #     # use only the features that are indexed in the target dataframes
-    #     self.df = self.df[self.df.index.isin(self.data_df.index)]
-    #     try:
-    #         # use only some features
-    #         selected_features = ast.literal_eval(
-    #             glob_conf.config["FEATS"]["praat.features"]
-    #         )
-    #         self.util.debug(
-    #             f"selecting features from Praat: {selected_features}"
-    #         )
-    #         sel_feats_df = pd.DataFrame()
-    #         hit = False
-    #         for feat in selected_features:
-    #             try:
-    #                 sel_feats_df[feat] = self.df[feat]
-    #                 hit = True
-    #             except KeyError:
-    #                 pass
-    #         if hit:
-    #             self.df = sel_feats_df
-    #             self.util.debug(
-    #                 "new feats shape after selecting Praat features:"
-    #                 f" {self.df.shape}"
-    #             )
-    #     except KeyError:
-    #         pass
+        tmp_audio_names = ["praat_audio_tmp.wav"]
+        audiofile.write(tmp_audio_names[0], signal, sr)
+        df = pd.DataFrame(index=tmp_audio_names)
+        index = audformat.utils.to_segmented_index(df.index, allow_nat=False)
+        df = feinberg_praat.compute_features(index)
+        df.set_index(index)
+        for i, col in enumerate(df.columns):
+            if df[col].isnull().values.any():
+                self.util.debug(
+                    f"{col} includes {df[col].isnull().sum()} nan,"
+                    " inserting mean values"
+                )
+                mean_val = df[col].mean()
+                if not np.isnan(mean_val):
+                    df[col] = df[col].fillna(mean_val)
+                else:
+                    df[col] = df[col].fillna(0)
+        df = df.astype(float)
+        feats = df.to_numpy()
+        return feats

nkululeko/feat_extract/feats_wav2vec2.py

@@ -19,7 +19,7 @@ class Wav2vec2(Featureset):
         cuda = "cuda" if torch.cuda.is_available() else "cpu"
         self.device = self.util.config_val("MODEL", "device", cuda)
         self.model_initialized = False
-        if feat_type == "wav2vec":
+        if feat_type == "wav2vec" or feat_type == "wav2vec2":
             self.feat_type = "wav2vec2-large-robust-ft-swbd-300h"
         else:
             self.feat_type = feat_type

nkululeko/feat_extract/feinberg_praat.py

@@ -199,28 +199,30 @@ def runPCA(df):
     # pickle.dump(x, f)
     # f.close()
 
-    x = StandardScaler().fit_transform(x)
-    if np.any(np.isnan(x)):
+    # x = StandardScaler().fit_transform(x)
+    if np.any(np.isnan(x[0])):
         print(
             f"Warning: {np.count_nonzero(np.isnan(x))} Nans in x, replacing" " with 0"
         )
         x[np.isnan(x)] = 0
-    if np.any(np.isfinite(x)):
-        print(f"Warning: {np.count_nonzero(np.isfinite(x))} infinite in x")
+    # if np.any(np.isfinite(x[0])):
+    #     print(f"Warning: {np.count_nonzero(np.isfinite(x))} finite in x")
 
     # PCA
     pca = PCA(n_components=2)
-    principalComponents = pca.fit_transform(x)
-    if np.any(np.isnan(principalComponents)):
-        print("pc is nan")
-        print(f"count: {np.count_nonzero(np.isnan(principalComponents))}")
-        print(principalComponents)
-        principalComponents = np.nan_to_num(principalComponents)
-
+    try:
+        principalComponents = pca.fit_transform(x)
+        if np.any(np.isnan(principalComponents)):
+            print("pc is nan")
+            print(f"count: {np.count_nonzero(np.isnan(principalComponents))}")
+            print(principalComponents)
+            principalComponents = np.nan_to_num(principalComponents)
+    except ValueError:
+        print("need more than one file for pca")
+        principalComponents = [[0, 0]]
     principalDf = pd.DataFrame(
         data=principalComponents, columns=["JitterPCA", "ShimmerPCA"]
     )
-
     return principalDf
 
 

nkululeko/models/model_mlp.py

@@ -174,6 +174,9 @@ class MLP_model(Model):
     def store(self):
         torch.save(self.model.state_dict(), self.store_path)
 
+    def store_as_onnx(self):
+        pass
+
     def load(self, run, epoch):
         self.set_id(run, epoch)
         dir = self.util.get_path("model_dir")

{nkululeko-0.74.2.dist-info → nkululeko-0.74.6.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: nkululeko
-Version: 0.74.2
+Version: 0.74.6
 Summary: Machine learning audio prediction experiments based on templates
 Home-page: https://github.com/felixbur/nkululeko
 Author: Felix Burkhardt
@@ -61,7 +61,7 @@ Requires-Dist: xgboost
 ## Overview
 A project to detect speaker characteristics by machine learning experiments with a high-level interface.
 
-The idea is to have a framework (based on e.g. sklearn and torch) that can be used to rapidly and automatically analyse and investigate audio data automatically.
+The idea is to have a framework (based on e.g. sklearn and torch) that can be used to rapidly and automatically analyse audio data and explore machine learning models based on that data.
 
 * NEW: Nkululeko now automatically generates PDF reports [sample for EmoDB](meta/images/emodb_report.pdf)
 * The latest features can be seen in [the ini-file](./ini_file.md) options that are used to control Nkululeko
@@ -308,6 +308,22 @@ F. Burkhardt, Johannes Wagner, Hagen Wierstorf, Florian Eyben and Björn Schulle
 Changelog
 =========
 
+Version 0.74.6
+--------------
+* added standard Wav2vec2 model
+
+Version 0.74.5
+--------------
+* added praat feature extractor for one sample
+
+Version 0.74.4
+--------------
+* fixed bug combining augmentations
+
+Version 0.74.3
+--------------
+* audiomentations interface changed
+
 Version 0.74.2
 --------------
 * combined augmentation methods

{nkululeko-0.74.2.dist-info → nkululeko-0.74.6.dist-info}/RECORD

@@ -1,11 +1,11 @@
 nkululeko/__init__.py,sha256=62f8HiEzJ8rG2QlTFJXUCMpvuH3fKI33DoJSj33mscc,63
-nkululeko/augment.py,sha256=NkP_OwC2ebUhcFnex0mD4KmcyawlPxiRmiaKy1WGiJc,2596
+nkululeko/augment.py,sha256=1kzUjscTPDrFxkR_HwnhPoB3SQQaKs5zQdWN9hRE6p4,2680
 nkululeko/balancer.py,sha256=WslJxQwMNnVYgZXF1y0ueS5zilRPQJZDhUG72Csb4Gw,11
 nkululeko/cacheddataset.py,sha256=lIJ6hUo5LoxSrzXtWV8mzwO7wRtUETWnOQ4ws2XfL1E,969
-nkululeko/constants.py,sha256=RdjwZCr5LNrLXdXn74oGOW3rZ2W5njgMIoso1tkFX-g,39
+nkululeko/constants.py,sha256=qEXgFJllipG3vjFOgo9g4RbcwzKbKqCzu42-olPPiT0,39
 nkululeko/demo.py,sha256=6CmLxH_0QJIMazPPg7IZur7ciNdqby6yOlh-6zu6YE0,1951
-nkululeko/demo_predictor.py,sha256=5PknI8SniGBRMzZOrmBOfPTV71rBbI1JCTnA6lXz6IU,2334
-nkululeko/experiment.py,sha256=tqcjHtl1zhut2WA9xjvIcg3B5NKlW4eF1k0KojofNk0,27924
+nkululeko/demo_predictor.py,sha256=z4t8IlwRsc-MrE83JlZ9KupOsW-Xalziu89nQD1FbCA,2623
+nkululeko/experiment.py,sha256=5vah4roe5PNc4jLSOGtpT5foVsO0n1y8GKUD371rHcE,28483
 nkululeko/explore.py,sha256=1OdBEYU5LYsuLTaW6WpDTciiGVpZQmZ-PYYiBd1HJtI,2251
 nkululeko/export.py,sha256=XqY7nFnta_hRFWeoqEwfCDz6BpCtPNNIs8r76o5g9rQ,4690
 nkululeko/feature_extractor.py,sha256=US5zFJ_DqReF9Q7Ynqo1qtamFPMYrgfs_I4VwIYSY8A,7275
@@ -27,7 +27,7 @@ nkululeko/test.py,sha256=Z00CQrJ6Pp9zycKSLrCFjzew-_AXll3pud2o0xur_KY,1457
 nkululeko/test_predictor.py,sha256=mO-jm1ViTtZY8QfWhJLyEboU1nn_CfKQ9c7-dgUxMp0,2403
 nkululeko/util.py,sha256=CY7vfFFa2XnWexq0HoIUIMxxwJ_JkATa2eQplyfqOX4,11293
 nkululeko/augmenting/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-nkululeko/augmenting/augmenter.py,sha256=czAWV320Kg8Em4M2gVeR8E0BRh46FPIJuAJSLSBqLR0,2408
+nkululeko/augmenting/augmenter.py,sha256=oodHgKO4DzHIGSryMLGQvDc9DYcQ6_XNajXSW813wNo,2300
 nkululeko/augmenting/randomsplicer.py,sha256=pgrTdwnd-I1CBbMx-do7QC5eJwx4z88bkqKzagl45OI,2389
 nkululeko/augmenting/randomsplicing.py,sha256=ldym9vZNsZIU5BAAaJVaOmAgmVHNs4a5i5K3bW-WAQU,1791
 nkululeko/augmenting/resampler.py,sha256=Bz-QMrcmH8eUgT_klIpQAgueVesPx72Erqr9fTFN5Ls,3413
@@ -59,16 +59,16 @@ nkululeko/feat_extract/feats_mld.py,sha256=RbRAaTTTfdIQeoDrGRsVUr5O-GVG443zbjdCy
 nkululeko/feat_extract/feats_mos.py,sha256=Bly7p6B0Guj4MQBdX_0G994lO5VUcmy5LLbXTSKi29Q,4247
 nkululeko/feat_extract/feats_opensmile.py,sha256=yDRGSiUQV3K3oLxVqq8Cxj5bkc-RiLzDYbAGKC9I5vc,4140
 nkululeko/feat_extract/feats_oxbow.py,sha256=N7uThvewVlH8HqSda-s_7UAtgXZkMwCOwUrsqeVHyLk,4830
-nkululeko/feat_extract/feats_praat.py,sha256=z0lhNY5CzBzH2-6REfYGcpv-yTSEozjGO6cWW_LX01Y,3366
+nkululeko/feat_extract/feats_praat.py,sha256=dqPhAUceze_6LN6vivXJFT2PPDXZKjmzYx0XnYKYGE8,3039
 nkululeko/feat_extract/feats_snr.py,sha256=zsxwRAzt3C-0B3EqT4p4zDaDVueMyyk8uxcaaa_ITso,2817
 nkululeko/feat_extract/feats_spectra.py,sha256=PLKoc_S3v3wibodUCiOnFFdF87U2rk2sfndRo2mmG64,3656
 nkululeko/feat_extract/feats_spkrec.py,sha256=VK4ma3uWzM0YZStsgRTirfkbzjWIfRWSgsYI038QlRY,4803
 nkululeko/feat_extract/feats_squim.py,sha256=8MaQ5lKfRqTJAub5VqEO9VziEVgMVTVe36CHkIQhGt4,4423
 nkululeko/feat_extract/feats_trill.py,sha256=vP7OKQCU8miz-NM9xEwP7kt-RL73uODcyqtUAkXM5Es,2994
-nkululeko/feat_extract/feats_wav2vec2.py,sha256=PmUCjVhP486KrgmwmXSAE6Pj2WMb0ADZBX-rA9zFoCY,4701
+nkululeko/feat_extract/feats_wav2vec2.py,sha256=r-HQ-oV6x9Ioe00gNMCTEI5iZuMyHvSUyKFDK_iXrdA,4728
 nkululeko/feat_extract/feats_wavlm.py,sha256=QoLQNYLFJ8BgEyx0lVgb48HHH9LYUOX7pJtKbgNLk1I,4509
 nkululeko/feat_extract/featureset.py,sha256=Xed_qbXFFasyKEyRpgCyZM3vYLqX0-O9RXnjN7hpUbY,1399
-nkululeko/feat_extract/feinberg_praat.py,sha256=bvy79CeKwWVIg55wrgcHwEEpnaFeX5fr9u4tnpaUIas,21095
+nkululeko/feat_extract/feinberg_praat.py,sha256=od8dV1ZRHytww70OwWK9Wm-M6nccOkT6CfaN3FJwRCY,21247
 nkululeko/losses/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 nkululeko/losses/loss_ccc.py,sha256=NOK0y0fxKUnU161B5geap6Fmn8QzoPl2MqtPiV8IuJE,976
 nkululeko/losses/loss_softf1loss.py,sha256=5gW-PuiqeAZcRgfwjueIOQtMokOjZWgQnVIv59HKTCo,1309
@@ -79,7 +79,7 @@ nkululeko/models/model_cnn.py,sha256=Omu7xPPeft3GwqZMySd2xF4IAm7WrzTAVjAnI_DsD6A
 nkululeko/models/model_gmm.py,sha256=onovzGBeguwZ-upXtuDLaBw9sd6fDDQslVBOrz1Z8TE,645
 nkululeko/models/model_knn.py,sha256=5tGqiPo2JTw9VLmD-MXNZKFJ5RTLA6uv_blJDJ9lScA,573
 nkululeko/models/model_knn_reg.py,sha256=Fbuk6Ku6eyrbbMEk7rB5dwfhvQOMsdZk6HI_0T0gYPw,580
-nkululeko/models/model_mlp.py,sha256=jAC5vPqAXwS1_ln2jk-LTr3jhv960tTLRtn1BZynNIs,8445
+nkululeko/models/model_mlp.py,sha256=YLbaC-4fDUeaozoUC4hT4oFlxyXfgkE3GtoR3wpp4Ho,8488
 nkululeko/models/model_mlp_regression.py,sha256=UXkMCHmLT-wl2aed8QmgJcvAebwRduUvdTLSzFsT1v4,9451
 nkululeko/models/model_svm.py,sha256=J1d8mf5T4QHtilkUTBkhegVB_0D2kRY0BiBGz-LUJmw,554
 nkululeko/models/model_svr.py,sha256=au5AtzjEpaY9_7Fz6CQoIZ3s6OAvLUxjAXuqHF9dRbk,514
@@ -103,8 +103,8 @@ nkululeko/split/split_utils.py,sha256=gdlXBwssgEogDBIbdVboeiYRWyJKYwuQm_-MkF1YCl
 nkululeko/utils/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 nkululeko/utils/files.py,sha256=UiGAtZRWYjHSvlmPaTMtzyNNGE6qaLaxQkybctS7iRM,4021
 nkululeko/utils/stats.py,sha256=zoZkrbELuukf9eKWh-EmzxKGjzJWQuCM18-2f_aIBz4,2554
-nkululeko-0.74.2.dist-info/LICENSE,sha256=0zGP5B_W35yAcGfHPS18Q2B8UhvLRY3dQq1MhpsJU_U,1076
-nkululeko-0.74.2.dist-info/METADATA,sha256=Y-wueiSa4AUvlb00GGdiyYmT22iiALAI5QU_zPrLvGs,29386
-nkululeko-0.74.2.dist-info/WHEEL,sha256=oiQVh_5PnQM0E3gPdiz09WCNmwiHDMaGer_elqB3coM,92
-nkululeko-0.74.2.dist-info/top_level.txt,sha256=DPFNNSHPjUeVKj44dVANAjuVGRCC3MusJ08lc2a8xFA,10
-nkululeko-0.74.2.dist-info/RECORD,,
+nkululeko-0.74.6.dist-info/LICENSE,sha256=0zGP5B_W35yAcGfHPS18Q2B8UhvLRY3dQq1MhpsJU_U,1076
+nkululeko-0.74.6.dist-info/METADATA,sha256=YbvU_3aAKddP86mjv373jJKNEOYKMTJl-AlfMD7E1xI,29686
+nkululeko-0.74.6.dist-info/WHEEL,sha256=oiQVh_5PnQM0E3gPdiz09WCNmwiHDMaGer_elqB3coM,92
+nkululeko-0.74.6.dist-info/top_level.txt,sha256=DPFNNSHPjUeVKj44dVANAjuVGRCC3MusJ08lc2a8xFA,10
+nkululeko-0.74.6.dist-info/RECORD,,