Shorten source file descriptions/headers

niklases · niklases · commit 10658b7e107d · 2025-07-11T19:08:15.000+02:00
diff --git a/pypef/__init__.py b/pypef/__init__.py
@@ -1,15 +1,4 @@
-#!/usr/bin/env python3
-# -*- coding: utf-8 -*-
-# Created on 05 October 2020
-# @authors: Niklas Siedhoff, Alexander-Maurice Illig
-# @contact: <niklas.siedhoff@rwth-aachen.de>
 # PyPEF - Pythonic Protein Engineering Framework
 # https://github.com/niklases/PyPEF
-# Licensed under Creative Commons Attribution-ShareAlike 4.0 International Public License (CC BY-SA 4.0)
-# For more information about the license see https://creativecommons.org/licenses/by-nc/4.0/legalcode
-
-# PyPEF – An Integrated Framework for Data-Driven Protein Engineering
-# Journal of Chemical Information and Modeling, 2021, 61, 3463-3476
-# https://doi.org/10.1021/acs.jcim.1c00099
 
 __version__ = '0.4.3-dev'
diff --git a/pypef/dca/gremlin_inference.py b/pypef/dca/gremlin_inference.py
@@ -1,16 +1,5 @@
-#!/usr/bin/env python3
-# -*- coding: utf-8 -*-
-# Created on 05 October 2020
-# @authors: Niklas Siedhoff, Alexander-Maurice Illig
-# @contact: <niklas.siedhoff@rwth-aachen.de>
 # PyPEF - Pythonic Protein Engineering Framework
 # https://github.com/niklases/PyPEF
-# Licensed under Creative Commons Attribution-ShareAlike 4.0 International Public License (CC BY-SA 4.0)
-# For more information about the license see https://creativecommons.org/licenses/by-nc/4.0/legalcode
-
-# PyPEF – An Integrated Framework for Data-Driven Protein Engineering
-# Journal of Chemical Information and Modeling, 2021, 61, 3463-3476
-# https://doi.org/10.1021/acs.jcim.1c00099
 
 """
 Code taken from GREMLIN repository available at https://github.com/sokrypton/GREMLIN_CPP/
diff --git a/pypef/dca/plmc_encoding.py b/pypef/dca/plmc_encoding.py
@@ -1,16 +1,5 @@
-#!/usr/bin/env python3
-# -*- coding: utf-8 -*-
-# Created on 05 October 2020
-# @authors: Niklas Siedhoff, Alexander-Maurice Illig
-# @contact: <niklas.siedhoff@rwth-aachen.de>
 # PyPEF - Pythonic Protein Engineering Framework
 # https://github.com/niklases/PyPEF
-# Licensed under Creative Commons Attribution-ShareAlike 4.0 International Public License (CC BY-SA 4.0)
-# For more information about the license see https://creativecommons.org/licenses/by-nc/4.0/legalcode
-
-# PyPEF – An Integrated Framework for Data-Driven Protein Engineering
-# Journal of Chemical Information and Modeling, 2021, 61, 3463-3476
-# https://doi.org/10.1021/acs.jcim.1c00099
 
 """
 Contains Python code used for the approach presented in our 'hybrid modeling' paper
diff --git a/pypef/gui/PyPEFGUIQtWindow.py b/pypef/gui/PyPEFGUIQtWindow.py
@@ -1,6 +1,8 @@
-# PyPEF Qt GUI window using PySide6
+# PyPEF - Pythonic Protein Engineering Framework
 # https://github.com/niklases/PyPEF
 
+# PyPEF Qt GUI window using PySide6
+
 import sys
 from os import getcwd, cpu_count, chdir
 
diff --git a/pypef/hybrid/hybrid_model.py b/pypef/hybrid/hybrid_model.py
@@ -1,16 +1,5 @@
-#!/usr/bin/env python3
-# -*- coding: utf-8 -*-
-# Created on 05 October 2020
-# @authors: Niklas Siedhoff, Alexander-Maurice Illig
-# @contact: <niklas.siedhoff@rwth-aachen.de>
 # PyPEF - Pythonic Protein Engineering Framework
 # https://github.com/niklases/PyPEF
-# Licensed under Creative Commons Attribution-ShareAlike 4.0 International Public License (CC BY-SA 4.0)
-# For more information about the license see https://creativecommons.org/licenses/by-nc/4.0/legalcode
-
-# PyPEF – An Integrated Framework for Data-Driven Protein Engineering
-# Journal of Chemical Information and Modeling, 2021, 61, 3463-3476
-# https://doi.org/10.1021/acs.jcim.1c00099
 
 # Contains Python code used for the approach presented in our 'hybrid modeling' paper
 # Preprint available at: https://doi.org/10.1101/2022.06.07.495081
diff --git a/pypef/hybrid/hybrid_run.py b/pypef/hybrid/hybrid_run.py
@@ -1,16 +1,5 @@
-#!/usr/bin/env python3
-# -*- coding: utf-8 -*-
-# Created on 05 October 2020
-# @authors: Niklas Siedhoff, Alexander-Maurice Illig
-# @contact: <niklas.siedhoff@rwth-aachen.de>
 # PyPEF - Pythonic Protein Engineering Framework
 # https://github.com/niklases/PyPEF
-# Licensed under Creative Commons Attribution-ShareAlike 4.0 International Public License (CC BY-SA 4.0)
-# For more information about the license see https://creativecommons.org/licenses/by-nc/4.0/legalcode
-
-# PyPEF – An Integrated Framework for Data-Driven Protein Engineering
-# Journal of Chemical Information and Modeling, 2021, 61, 3463-3476
-# https://doi.org/10.1021/acs.jcim.1c00099
 
 import logging
 logger = logging.getLogger('pypef.dca.dca_run')
diff --git a/pypef/llm/esm_lora_tune.py b/pypef/llm/esm_lora_tune.py
@@ -1,5 +1,5 @@
-# Niklas Siedhoff
 # PyPEF - Pythonic Protein Engineering Framework
+# https://github.com/niklases/PyPEF
 
 # Using (training, testing/infering) ESM model(s) (e.g. ESM1v) published under 
 # MIT License
diff --git a/pypef/llm/prosst_lora_tune.py b/pypef/llm/prosst_lora_tune.py
@@ -1,5 +1,5 @@
-# Niklas Siedhoff
 # PyPEF - Pythonic Protein Engineering Framework
+# https://github.com/niklases/PyPEF
 
 # Using (training, testing/infering) ProSST model(s) published under 
 # GNU GENERAL PUBLIC LICENSE: GPL-3.0 license
diff --git a/pypef/llm/prosst_structure/encoder/gvp.py b/pypef/llm/prosst_structure/encoder/gvp.py
@@ -1,6 +1,5 @@
-
-# Niklas Siedhoff
 # PyPEF - Pythonic Protein Engineering Framework
+# https://github.com/niklases/PyPEF
 
 # Using (training, testing/infering) ProSST model(s) published under 
 # GNU GENERAL PUBLIC LICENSE: GPL-3.0 license
diff --git a/pypef/llm/prosst_structure/encoder/layer.py b/pypef/llm/prosst_structure/encoder/layer.py
@@ -1,6 +1,5 @@
-
-# Niklas Siedhoff
 # PyPEF - Pythonic Protein Engineering Framework
+# https://github.com/niklases/PyPEF
 
 # Using (training, testing/infering) ProSST model(s) published under 
 # GNU GENERAL PUBLIC LICENSE: GPL-3.0 license
diff --git a/pypef/llm/prosst_structure/quantizer.py b/pypef/llm/prosst_structure/quantizer.py
@@ -1,6 +1,5 @@
-
-# Niklas Siedhoff
 # PyPEF - Pythonic Protein Engineering Framework
+# https://github.com/niklases/PyPEF
 
 # Using (training, testing/infering) ProSST model(s) published under 
 # GNU GENERAL PUBLIC LICENSE: GPL-3.0 license
diff --git a/pypef/llm/prosst_structure/scatter.py b/pypef/llm/prosst_structure/scatter.py
@@ -1,7 +1,5 @@
-
-
-# Niklas Siedhoff
 # PyPEF - Pythonic Protein Engineering Framework
+# https://github.com/niklases/PyPEF
 
 # Code taken from PyTorch Scatter
 # https://github.com/rusty1s/pytorch_scatter
diff --git a/pypef/main.py b/pypef/main.py
@@ -1,16 +1,5 @@
-#!/usr/bin/env python3
-# -*- coding: utf-8 -*-
-# Created on 05 October 2020
-# @authors: Niklas Siedhoff, Alexander-Maurice Illig
-# @contact: <niklas.siedhoff@rwth-aachen.de>
 # PyPEF - Pythonic Protein Engineering Framework
 # https://github.com/niklases/PyPEF
-# Licensed under Creative Commons Attribution-ShareAlike 4.0 International Public License (CC BY-SA 4.0)
-# For more information about the license see https://creativecommons.org/licenses/by-nc/4.0/legalcode
-
-# PyPEF – An Integrated Framework for Data-Driven Protein Engineering
-# Journal of Chemical Information and Modeling, 2021, 61, 3463-3476
-# https://doi.org/10.1021/acs.jcim.1c00099
 
 # docstring for argument parsing using docopts
 """
diff --git a/pypef/ml/ml_run.py b/pypef/ml/ml_run.py
@@ -1,16 +1,5 @@
-#!/usr/bin/env python3
-# -*- coding: utf-8 -*-
-# Created on 05 October 2020
-# @authors: Niklas Siedhoff, Alexander-Maurice Illig
-# @contact: <niklas.siedhoff@rwth-aachen.de>
 # PyPEF - Pythonic Protein Engineering Framework
 # https://github.com/niklases/PyPEF
-# Licensed under Creative Commons Attribution-ShareAlike 4.0 International Public License (CC BY-SA 4.0)
-# For more information about the license see https://creativecommons.org/licenses/by-nc/4.0/legalcode
-
-# PyPEF – An Integrated Framework for Data-Driven Protein Engineering
-# Journal of Chemical Information and Modeling, 2021, 61, 3463-3476
-# https://doi.org/10.1021/acs.jcim.1c00099
 
 import os
 from os import listdir
diff --git a/pypef/ml/parallelization.py b/pypef/ml/parallelization.py
@@ -1,16 +1,5 @@
-#!/usr/bin/env python3
-# -*- coding: utf-8 -*-
-# Created on 05 October 2020
-# @authors: Niklas Siedhoff, Alexander-Maurice Illig
-# @contact: <niklas.siedhoff@rwth-aachen.de>
 # PyPEF - Pythonic Protein Engineering Framework
 # https://github.com/niklases/PyPEF
-# Licensed under Creative Commons Attribution-ShareAlike 4.0 International Public License (CC BY-SA 4.0)
-# For more information about the license see https://creativecommons.org/licenses/by-nc/4.0/legalcode
-
-# PyPEF – An Integrated Framework for Data-Driven Protein Engineering
-# Journal of Chemical Information and Modeling, 2021, 61, 3463-3476
-# https://doi.org/10.1021/acs.jcim.1c00099
 
 """
 Validation modules from regression.py for AAindex-based encoding
diff --git a/pypef/ml/regression.py b/pypef/ml/regression.py
@@ -1,16 +1,5 @@
-#!/usr/bin/env python3
-# -*- coding: utf-8 -*-
-# Created on 05 October 2020
-# @authors: Niklas Siedhoff, Alexander-Maurice Illig
-# @contact: <niklas.siedhoff@rwth-aachen.de>
 # PyPEF - Pythonic Protein Engineering Framework
 # https://github.com/niklases/PyPEF
-# Licensed under Creative Commons Attribution-ShareAlike 4.0 International Public License (CC BY-SA 4.0)
-# For more information about the license see https://creativecommons.org/licenses/by-nc/4.0/legalcode
-
-# PyPEF – An Integrated Framework for Data-Driven Protein Engineering
-# Journal of Chemical Information and Modeling, 2021, 61, 3463-3476
-# https://doi.org/10.1021/acs.jcim.1c00099
 
 """
 Main modules for regression/ML including feature generation
diff --git a/pypef/settings.py b/pypef/settings.py
@@ -1,16 +1,4 @@
-#!/usr/bin/env python3
-# -*- coding: utf-8 -*-
-# Created on 05 October 2020
-# @authors: Niklas Siedhoff, Alexander-Maurice Illig
-# @contact: <niklas.siedhoff@rwth-aachen.de>
 # PyPEF - Pythonic Protein Engineering Framework
 # https://github.com/niklases/PyPEF
-# Licensed under Creative Commons Attribution-ShareAlike 4.0 International Public License (CC BY-SA 4.0)
-# For more information about the license see https://creativecommons.org/licenses/by-nc/4.0/legalcode
-
-# PyPEF – An Integrated Framework for Data-Driven Protein Engineering
-# Journal of Chemical Information and Modeling, 2021, 61, 3463-3476
-# https://doi.org/10.1021/acs.jcim.1c00099
-
 
 USE_RAY = False
diff --git a/pypef/utils/directed_evolution.py b/pypef/utils/directed_evolution.py
@@ -1,16 +1,5 @@
-#!/usr/bin/env python3
-# -*- coding: utf-8 -*-
-# Created on 05 October 2020
-# @authors: Niklas Siedhoff, Alexander-Maurice Illig
-# @contact: <niklas.siedhoff@rwth-aachen.de>
 # PyPEF - Pythonic Protein Engineering Framework
 # https://github.com/niklases/PyPEF
-# Licensed under Creative Commons Attribution-ShareAlike 4.0 International Public License (CC BY-SA 4.0)
-# For more information about the license see https://creativecommons.org/licenses/by-nc/4.0/legalcode
-
-# PyPEF – An Integrated Framework for Data-Driven Protein Engineering
-# Journal of Chemical Information and Modeling, 2021, 61, 3463-3476
-# https://doi.org/10.1021/acs.jcim.1c00099
 
 """
 Modules for performing random evolution walks
diff --git a/pypef/utils/helpers.py b/pypef/utils/helpers.py
@@ -1,3 +1,5 @@
+# PyPEF - Pythonic Protein Engineering Framework
+# https://github.com/niklases/PyPEF
 
 import shutil
 import subprocess
diff --git a/pypef/utils/learning_test_sets.py b/pypef/utils/learning_test_sets.py
@@ -1,16 +1,5 @@
-#!/usr/bin/env python3
-# -*- coding: utf-8 -*-
-# Created on 05 October 2020
-# @authors: Niklas Siedhoff, Alexander-Maurice Illig
-# @contact: <niklas.siedhoff@rwth-aachen.de>
 # PyPEF - Pythonic Protein Engineering Framework
 # https://github.com/niklases/PyPEF
-# Licensed under Creative Commons Attribution-ShareAlike 4.0 International Public License (CC BY-SA 4.0)
-# For more information about the license see https://creativecommons.org/licenses/by-nc/4.0/legalcode
-
-# PyPEF – An Integrated Framework for Data-Driven Protein Engineering
-# Journal of Chemical Information and Modeling, 2021, 61, 3463-3476
-# https://doi.org/10.1021/acs.jcim.1c00099
 
 """
 Modules for creating training and test sets
diff --git a/pypef/utils/low_n_mutation_extrapolation.py b/pypef/utils/low_n_mutation_extrapolation.py
@@ -1,16 +1,5 @@
-#!/usr/bin/env python3
-# -*- coding: utf-8 -*-
-# Created on 05 October 2020
-# @authors: Niklas Siedhoff, Alexander-Maurice Illig
-# @contact: <niklas.siedhoff@rwth-aachen.de>
 # PyPEF - Pythonic Protein Engineering Framework
 # https://github.com/niklases/PyPEF
-# Licensed under Creative Commons Attribution-ShareAlike 4.0 International Public License (CC BY-SA 4.0)
-# For more information about the license see https://creativecommons.org/licenses/by-nc/4.0/legalcode
-
-# PyPEF – An Integrated Framework for Data-Driven Protein Engineering
-# Journal of Chemical Information and Modeling, 2021, 61, 3463-3476
-# https://doi.org/10.1021/acs.jcim.1c00099
 
 import os
 import random
diff --git a/pypef/utils/performance.py b/pypef/utils/performance.py
@@ -1,16 +1,5 @@
-#!/usr/bin/env python3
-# -*- coding: utf-8 -*-
-# Created on 05 October 2020
-# @authors: Niklas Siedhoff, Alexander-Maurice Illig
-# @contact: <niklas.siedhoff@rwth-aachen.de>
 # PyPEF - Pythonic Protein Engineering Framework
 # https://github.com/niklases/PyPEF
-# Licensed under Creative Commons Attribution-ShareAlike 4.0 International Public License (CC BY-SA 4.0)
-# For more information about the license see https://creativecommons.org/licenses/by-nc/4.0/legalcode
-
-# PyPEF – An Integrated Framework for Data-Driven Protein Engineering
-# Journal of Chemical Information and Modeling, 2021, 61, 3463-3476
-# https://doi.org/10.1021/acs.jcim.1c00099
 
 import warnings
 import numpy as np
diff --git a/pypef/utils/plot.py b/pypef/utils/plot.py
@@ -1,16 +1,5 @@
-#!/usr/bin/env python3
-# -*- coding: utf-8 -*-
-# Created on 05 October 2020
-# @authors: Niklas Siedhoff, Alexander-Maurice Illig
-# @contact: <niklas.siedhoff@rwth-aachen.de>
 # PyPEF - Pythonic Protein Engineering Framework
 # https://github.com/niklases/PyPEF
-# Licensed under Creative Commons Attribution-ShareAlike 4.0 International Public License (CC BY-SA 4.0)
-# For more information about the license see https://creativecommons.org/licenses/by-nc/4.0/legalcode
-
-# PyPEF – An Integrated Framework for Data-Driven Protein Engineering
-# Journal of Chemical Information and Modeling, 2021, 61, 3463-3476
-# https://doi.org/10.1021/acs.jcim.1c00099
 
 import os
 import numpy as np
diff --git a/pypef/utils/prediction_sets.py b/pypef/utils/prediction_sets.py
@@ -1,16 +1,5 @@
-#!/usr/bin/env python3
-# -*- coding: utf-8 -*-
-# Created on 05 October 2020
-# @authors: Niklas Siedhoff, Alexander-Maurice Illig
-# @contact: <niklas.siedhoff@rwth-aachen.de>
 # PyPEF - Pythonic Protein Engineering Framework
 # https://github.com/niklases/PyPEF
-# Licensed under Creative Commons Attribution-ShareAlike 4.0 International Public License (CC BY-SA 4.0)
-# For more information about the license see https://creativecommons.org/licenses/by-nc/4.0/legalcode
-
-# PyPEF – An Integrated Framework for Data-Driven Protein Engineering
-# Journal of Chemical Information and Modeling, 2021, 61, 3463-3476
-# https://doi.org/10.1021/acs.jcim.1c00099
 
 """
 Modules for making prediction files
diff --git a/pypef/utils/split.py b/pypef/utils/split.py
@@ -1,3 +1,5 @@
+# PyPEF - Pythonic Protein Engineering Framework
+# https://github.com/niklases/PyPEF
 
 
 import pandas as pd
@@ -6,6 +8,9 @@
 import matplotlib.pyplot as plt
 from sklearn.model_selection import KFold
 
+import logging
+logger = logging.getLogger('pypef.utils.split')
+
 
 class DatasetSplitter:
     def __init__(
@@ -109,15 +114,15 @@ def print_shapes(self):
         """
         random_shape_train = [np.shape(k) for k in self.random_splits_train_indices_combined]
         random_shape_test = [np.shape(k) for k in self.random_splits_test_indices_combined]
-        print(f'Random train --> test split shapes: {random_shape_train} --> {random_shape_test}')
+        logger.info(f'Random train --> test split shapes: {random_shape_train} --> {random_shape_test}')
         
         modulo_shape_train = [np.shape(k) for k in self.modulo_splits_train_indices_combined]
         modulo_shape_test = [np.shape(k) for k in self.modulo_splits_test_indices_combined]
-        print(f'Modulo train --> test split shapes: {modulo_shape_train} --> {modulo_shape_test}')
+        logger.info(f'Modulo train --> test split shapes: {modulo_shape_train} --> {modulo_shape_test}')
         
         cont_shape_train = [np.shape(k) for k in self.cont_splits_train_indices_combined]
         cont_shape_test = [np.shape(k) for k in self.cont_splits_test_indices_combined]
-        print(f'Continuous train --> test split shapes: {cont_shape_train} --> {cont_shape_test}')
+        logger.info(f'Continuous train --> test split shapes: {cont_shape_train} --> {cont_shape_test}')
 
     def _get_zero_counts(self) -> dict:
         all_poses = np.asarray(range(self.min_pos, self.max_pos + 1))
@@ -187,11 +192,5 @@ def plot_distributions(self):
         axs[0, self.n_cv // 2].set_xticks(xticks)
         fig_path = path.abspath(path.splitext(path.basename(self.csv_file))[0] + '_pos_aa_distr.png')
         plt.savefig(fig_path, dpi=300)
-        print(f"Saved figure as {fig_path}")
+        logger.info(f"Saved figure as {fig_path}.")
         plt.close(fig)
-
-
-if __name__ == '__main__':
-    d = DatasetSplitter('C:\dev\DMS_ProteinGym_substitutions\BLAT_ECOLX_Stiffler_2015.csv')
-    d.print_shapes()
-    d.plot_distributions()
diff --git a/pypef/utils/sto2a2m.py b/pypef/utils/sto2a2m.py
@@ -1,16 +1,5 @@
-#!/usr/bin/env python3
-# -*- coding: utf-8 -*-
-# Created on 05 October 2020
-# @authors: Niklas Siedhoff, Alexander-Maurice Illig
-# @contact: <niklas.siedhoff@rwth-aachen.de>
 # PyPEF - Pythonic Protein Engineering Framework
 # https://github.com/niklases/PyPEF
-# Licensed under Creative Commons Attribution-ShareAlike 4.0 International Public License (CC BY-SA 4.0)
-# For more information about the license see https://creativecommons.org/licenses/by-nc/4.0/legalcode
-
-# PyPEF – An Integrated Framework for Data-Driven Protein Engineering
-# Journal of Chemical Information and Modeling, 2021, 61, 3463-3476
-# https://doi.org/10.1021/acs.jcim.1c00099
 
 import logging
 logger = logging.getLogger('pypef.utils.sto2a2m')
diff --git a/pypef/utils/to_file.py b/pypef/utils/to_file.py
diff --git a/pypef/utils/utils_run.py b/pypef/utils/utils_run.py
diff --git a/pypef/utils/variant_data.py b/pypef/utils/variant_data.py
