nkululeko 0.86.4__tar.gz → 0.86.5__tar.gz

{nkululeko-0.86.4 → nkululeko-0.86.5}/CHANGELOG.md

@@ -1,6 +1,10 @@
 Changelog
 =========
 
+Version 0.86.5
+--------------
+* fix audio path detection in data csv import
+
 Version 0.86.4
 --------------
 * add finetuning to the demo module

{nkululeko-0.86.4/nkululeko.egg-info → nkululeko-0.86.5}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: nkululeko
-Version: 0.86.4
+Version: 0.86.5
 Summary: Machine learning audio prediction experiments based on templates
 Home-page: https://github.com/felixbur/nkululeko
 Author: Felix Burkhardt
@@ -58,6 +58,8 @@ Requires-Dist: pylatex
   - [Hello World example](#hello-world-example)
   - [Features](#features)
 - [License](#license)
+- [Contributing](#contributing)
+- [Citing](#citing)
 
  
 ## Overview
@@ -65,7 +67,7 @@ A project to detect speaker characteristics by machine learning experiments with
 
 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)
+* NEW with nkululeko: [Finetune transformer-models](http://blog.syntheticspeech.de/2024/05/29/nkululeko-how-to-finetune-a-transformer-model/)
 * The latest features can be seen in [the ini-file](./ini_file.md) options that are used to control Nkululeko
 * Below is a [Hello World example](#helloworld) that should set you up fastly, also on [Google Colab](https://colab.research.google.com/drive/1GYNBd5cdZQ1QC3Jm58qoeMaJg3UuPhjw?usp=sharing#scrollTo=4G_SjuF9xeQf), and [with Kaggle](https://www.kaggle.com/felixburk/nkululeko-hello-world-example)
 * [Here's a blog post on how to set up nkululeko on your computer.](http://blog.syntheticspeech.de/2021/08/30/how-to-set-up-your-first-nkululeko-project/)
@@ -249,7 +251,7 @@ There's my [blog](http://blog.syntheticspeech.de/?s=nkululeko) with tutorials:
 * [Predict new labels for your data from public models and check bias](http://blog.syntheticspeech.de/2023/08/16/nkululeko-how-to-predict-labels-for-your-data-from-existing-models-and-check-them/)
 * [Resample](http://blog.syntheticspeech.de/2023/08/31/how-to-fix-different-sampling-rates-in-a-dataset-with-nkululeko/)
 * [Get some statistics on correlation and effect-size](http://blog.syntheticspeech.de/2023/09/05/nkululeko-get-some-statistics-on-correlation-and-effect-size/)
-* [Generate a latex / pdf report](http://blog.syntheticspeech.de/2023/09/26/nkululeko-generate-a-latex-pdf-report/) 
+* [Automatic generation  of a latex / pdf report](http://blog.syntheticspeech.de/2023/09/26/nkululeko-generate-a-latex-pdf-report/) 
 * [Inspect your data with Spotlight](http://blog.syntheticspeech.de/2023/10/31/nkululeko-inspect-your-data-with-spotlight/)
 * [Automatically stratify your split sets](http://blog.syntheticspeech.de/2023/11/07/nkululeko-automatically-stratify-your-split-sets/)
 * [re-name data column names](http://blog.syntheticspeech.de/2023/11/16/nkululeko-re-name-data-column-names/)
@@ -314,6 +316,12 @@ Here's [an animation that shows the progress of classification done with nkulule
 
 ## License
 Nkululeko can be used under the [MIT license](https://choosealicense.com/licenses/mit/)
+
+
+## Contributing
+Contributions are welcome and encouraged. To learn more about how to contribute to nkululeko please refer to the [Contributing guidelines](./CONTRIBUTING.md)
+
+## Citing
 If you use it, please mention the Nkululeko paper
 
 F. Burkhardt, Johannes Wagner, Hagen Wierstorf, Florian Eyben and Björn Schuller: Nkululeko: A Tool For Rapid Speaker Characteristics Detection, Proc. Proc. LREC, 2022
@@ -335,6 +343,10 @@ F. Burkhardt, Johannes Wagner, Hagen Wierstorf, Florian Eyben and Björn Schulle
 Changelog
 =========
 
+Version 0.86.5
+--------------
+* fix audio path detection in data csv import
+
 Version 0.86.4
 --------------
 * add finetuning to the demo module

{nkululeko-0.86.4 → nkululeko-0.86.5}/README.md

@@ -14,6 +14,8 @@
   - [Hello World example](#hello-world-example)
   - [Features](#features)
 - [License](#license)
+- [Contributing](#contributing)
+- [Citing](#citing)
 
  
 ## Overview
@@ -21,7 +23,7 @@ A project to detect speaker characteristics by machine learning experiments with
 
 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)
+* NEW with nkululeko: [Finetune transformer-models](http://blog.syntheticspeech.de/2024/05/29/nkululeko-how-to-finetune-a-transformer-model/)
 * The latest features can be seen in [the ini-file](./ini_file.md) options that are used to control Nkululeko
 * Below is a [Hello World example](#helloworld) that should set you up fastly, also on [Google Colab](https://colab.research.google.com/drive/1GYNBd5cdZQ1QC3Jm58qoeMaJg3UuPhjw?usp=sharing#scrollTo=4G_SjuF9xeQf), and [with Kaggle](https://www.kaggle.com/felixburk/nkululeko-hello-world-example)
 * [Here's a blog post on how to set up nkululeko on your computer.](http://blog.syntheticspeech.de/2021/08/30/how-to-set-up-your-first-nkululeko-project/)
@@ -205,7 +207,7 @@ There's my [blog](http://blog.syntheticspeech.de/?s=nkululeko) with tutorials:
 * [Predict new labels for your data from public models and check bias](http://blog.syntheticspeech.de/2023/08/16/nkululeko-how-to-predict-labels-for-your-data-from-existing-models-and-check-them/)
 * [Resample](http://blog.syntheticspeech.de/2023/08/31/how-to-fix-different-sampling-rates-in-a-dataset-with-nkululeko/)
 * [Get some statistics on correlation and effect-size](http://blog.syntheticspeech.de/2023/09/05/nkululeko-get-some-statistics-on-correlation-and-effect-size/)
-* [Generate a latex / pdf report](http://blog.syntheticspeech.de/2023/09/26/nkululeko-generate-a-latex-pdf-report/) 
+* [Automatic generation  of a latex / pdf report](http://blog.syntheticspeech.de/2023/09/26/nkululeko-generate-a-latex-pdf-report/) 
 * [Inspect your data with Spotlight](http://blog.syntheticspeech.de/2023/10/31/nkululeko-inspect-your-data-with-spotlight/)
 * [Automatically stratify your split sets](http://blog.syntheticspeech.de/2023/11/07/nkululeko-automatically-stratify-your-split-sets/)
 * [re-name data column names](http://blog.syntheticspeech.de/2023/11/16/nkululeko-re-name-data-column-names/)
@@ -270,6 +272,12 @@ Here's [an animation that shows the progress of classification done with nkulule
 
 ## License
 Nkululeko can be used under the [MIT license](https://choosealicense.com/licenses/mit/)
+
+
+## Contributing
+Contributions are welcome and encouraged. To learn more about how to contribute to nkululeko please refer to the [Contributing guidelines](./CONTRIBUTING.md)
+
+## Citing
 If you use it, please mention the Nkululeko paper
 
 F. Burkhardt, Johannes Wagner, Hagen Wierstorf, Florian Eyben and Björn Schuller: Nkululeko: A Tool For Rapid Speaker Characteristics Detection, Proc. Proc. LREC, 2022

{nkululeko-0.86.4 → nkululeko-0.86.5}/nkululeko/constants.py

@@ -1,2 +1,2 @@
-VERSION="0.86.4"
+VERSION="0.86.5"
 SAMPLING_RATE = 16000

{nkululeko-0.86.4 → nkululeko-0.86.5}/nkululeko/data/dataset_csv.py

@@ -59,6 +59,20 @@ class Dataset_CSV(Dataset):
                         lambda x: root + "/" + audio_path + "/" + x
                     )
                 )
+        else: # absolute path is True
+            if audformat.index_type(df.index) == "segmented":
+                file_index = (
+                    df.index.levels[0]
+                    .map(lambda x: audio_path + "/" + x)
+                    .values
+                )
+                df = df.set_index(df.index.set_levels(
+                    file_index, level="file"))
+            else:
+                if not isinstance(df, pd.DataFrame):
+                    df = pd.DataFrame(df)
+                df = df.set_index(df.index.to_series().apply(
+                    lambda x: audio_path + "/" + x ))
 
         self.df = df
         self.db = None

{nkululeko-0.86.4 → nkululeko-0.86.5}/nkululeko/experiment.py

@@ -112,11 +112,11 @@ class Experiment:
         auto_labels = list(next(iter(self.datasets.values())).df[self.target].unique())
         if labels:
             self.labels = ast.literal_eval(labels)
-            self.util.debug(f"Target labels (from config): {labels}")
+            self.util.debug(f"Using target labels (from config): {labels}")
         else:
             self.labels = auto_labels
         # print autolabel no matter it is specified or not
-        self.util.debug(f"Target labels (from database): {auto_labels}")
+        self.util.debug(f"Labels (from database): {auto_labels}")
         glob_conf.set_labels(self.labels)
         self.util.debug(f"loaded databases {dbs}")
 

{nkululeko-0.86.4 → nkululeko-0.86.5}/nkululeko/reporting/reporter.py

@@ -7,10 +7,11 @@ from confidence_intervals import evaluate_with_conf_int
 import matplotlib.pyplot as plt
 import numpy as np
 from scipy.stats import pearsonr
-from sklearn.metrics import ConfusionMatrixDisplay
+from sklearn.metrics import ConfusionMatrixDisplay, roc_curve
 from sklearn.metrics import classification_report
 from sklearn.metrics import confusion_matrix
 from sklearn.metrics import r2_score
+from sklearn.metrics import roc_curve, auc, roc_auc_score
 from torch import is_tensor
 
 from audmetric import accuracy
@@ -262,8 +263,15 @@ class Reporter:
                     c_ress[i] = float(f"{c_res:.3f}")
                 self.util.debug(f"labels: {labels}")
                 f1_per_class = f"result per class (F1 score): {c_ress}"
+                if len(np.unique(self.truths)) == 2:
+                    fpr, tpr, _ = roc_curve(self.truths, self.preds)
+                    auc_score = auc(fpr, tpr)
+                    pauc_score = roc_auc_score(self.truths, self.preds, max_fpr=0.1)
+                    auc_pauc = f"auc: {auc_score:.3f}, pauc: {pauc_score:.3f}"
+                    self.util.debug(auc_pauc)
                 self.util.debug(f1_per_class)
                 rpt_str = f"{json.dumps(rpt)}\n{f1_per_class}"
+                # rpt_str += f"\n{auc_auc}"
                 text_file.write(rpt_str)
                 glob_conf.report.add_item(
                     ReportItem(

{nkululeko-0.86.4 → nkululeko-0.86.5/nkululeko.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: nkululeko
-Version: 0.86.4
+Version: 0.86.5
 Summary: Machine learning audio prediction experiments based on templates
 Home-page: https://github.com/felixbur/nkululeko
 Author: Felix Burkhardt
@@ -58,6 +58,8 @@ Requires-Dist: pylatex
   - [Hello World example](#hello-world-example)
   - [Features](#features)
 - [License](#license)
+- [Contributing](#contributing)
+- [Citing](#citing)
 
  
 ## Overview
@@ -65,7 +67,7 @@ A project to detect speaker characteristics by machine learning experiments with
 
 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)
+* NEW with nkululeko: [Finetune transformer-models](http://blog.syntheticspeech.de/2024/05/29/nkululeko-how-to-finetune-a-transformer-model/)
 * The latest features can be seen in [the ini-file](./ini_file.md) options that are used to control Nkululeko
 * Below is a [Hello World example](#helloworld) that should set you up fastly, also on [Google Colab](https://colab.research.google.com/drive/1GYNBd5cdZQ1QC3Jm58qoeMaJg3UuPhjw?usp=sharing#scrollTo=4G_SjuF9xeQf), and [with Kaggle](https://www.kaggle.com/felixburk/nkululeko-hello-world-example)
 * [Here's a blog post on how to set up nkululeko on your computer.](http://blog.syntheticspeech.de/2021/08/30/how-to-set-up-your-first-nkululeko-project/)
@@ -249,7 +251,7 @@ There's my [blog](http://blog.syntheticspeech.de/?s=nkululeko) with tutorials:
 * [Predict new labels for your data from public models and check bias](http://blog.syntheticspeech.de/2023/08/16/nkululeko-how-to-predict-labels-for-your-data-from-existing-models-and-check-them/)
 * [Resample](http://blog.syntheticspeech.de/2023/08/31/how-to-fix-different-sampling-rates-in-a-dataset-with-nkululeko/)
 * [Get some statistics on correlation and effect-size](http://blog.syntheticspeech.de/2023/09/05/nkululeko-get-some-statistics-on-correlation-and-effect-size/)
-* [Generate a latex / pdf report](http://blog.syntheticspeech.de/2023/09/26/nkululeko-generate-a-latex-pdf-report/) 
+* [Automatic generation  of a latex / pdf report](http://blog.syntheticspeech.de/2023/09/26/nkululeko-generate-a-latex-pdf-report/) 
 * [Inspect your data with Spotlight](http://blog.syntheticspeech.de/2023/10/31/nkululeko-inspect-your-data-with-spotlight/)
 * [Automatically stratify your split sets](http://blog.syntheticspeech.de/2023/11/07/nkululeko-automatically-stratify-your-split-sets/)
 * [re-name data column names](http://blog.syntheticspeech.de/2023/11/16/nkululeko-re-name-data-column-names/)
@@ -314,6 +316,12 @@ Here's [an animation that shows the progress of classification done with nkulule
 
 ## License
 Nkululeko can be used under the [MIT license](https://choosealicense.com/licenses/mit/)
+
+
+## Contributing
+Contributions are welcome and encouraged. To learn more about how to contribute to nkululeko please refer to the [Contributing guidelines](./CONTRIBUTING.md)
+
+## Citing
 If you use it, please mention the Nkululeko paper
 
 F. Burkhardt, Johannes Wagner, Hagen Wierstorf, Florian Eyben and Björn Schuller: Nkululeko: A Tool For Rapid Speaker Characteristics Detection, Proc. Proc. LREC, 2022
@@ -335,6 +343,10 @@ F. Burkhardt, Johannes Wagner, Hagen Wierstorf, Florian Eyben and Björn Schulle
 Changelog
 =========
 
+Version 0.86.5
+--------------
+* fix audio path detection in data csv import
+
 Version 0.86.4
 --------------
 * add finetuning to the demo module