extrap 4.2.2__tar.gz → 4.2.4__tar.gz

{extrap-4.2.2/extrap.egg-info → extrap-4.2.4}/PKG-INFO

@@ -1,6 +1,6 @@
-Metadata-Version: 2.1
+Metadata-Version: 2.4
 Name: extrap
-Version: 4.2.2
+Version: 4.2.4
 Summary: Extra-P, automated performance modeling for HPC applications
 Home-page: https://github.com/extra-p/extrap
 Author: Extra-P project
@@ -15,6 +15,8 @@ Classifier: Programming Language :: Python :: 3.9
 Classifier: Programming Language :: Python :: 3.10
 Classifier: Programming Language :: Python :: 3.11
 Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
+Classifier: Programming Language :: Python :: 3.14
 Classifier: Operating System :: OS Independent
 Classifier: Intended Audience :: Developers
 Classifier: Intended Audience :: Science/Research
@@ -36,6 +38,17 @@ Requires-Dist: pyobjc-framework-Cocoa~=9.0; sys_platform == "darwin"
 Requires-Dist: scikit-learn>=1.2.2
 Provides-Extra: adaptive-modeling
 Requires-Dist: extrap-adaptive-modeler<3,>=1.0.2; extra == "adaptive-modeling"
+Dynamic: author
+Dynamic: author-email
+Dynamic: classifier
+Dynamic: description
+Dynamic: description-content-type
+Dynamic: home-page
+Dynamic: license-file
+Dynamic: provides-extra
+Dynamic: requires-dist
+Dynamic: requires-python
+Dynamic: summary
 
 # Extra-P
 
@@ -154,31 +167,51 @@ You can find an overview about all command line options under [docs/command-line
 
 Please cite Extra-P in your publications if it helps your research:
 
-    @inproceedings{calotoiu_ea:2013:modeling,
-      author = {Calotoiu, Alexandru and Hoefler, Torsten and Poke, Marius and Wolf, Felix},
-      month = {November},
-      title = {Using Automated Performance Modeling to Find Scalability Bugs in Complex Codes},
-      booktitle = {Proc. of the ACM/IEEE Conference on Supercomputing (SC13), Denver, CO, USA},
-      year = {2013},
-      pages = {1--12},
-      publisher = {ACM},
-      isbn = {978-1-4503-2378-9},
-      doi = {10.1145/2503210.2503277}
-    }
-
-### Publications
-
-1. Alexandru Calotoiu, David Beckingsale, Christopher W. Earl, Torsten Hoefler, Ian Karlin, Martin Schulz, Felix Wolf:
-   Fast Multi-Parameter Performance Modeling. In Proc. of the 2016 IEEE International Conference on Cluster Computing (
+Alexandru Calotoiu, Marcin Copik, Fabian Czappa, Alexander Geiß, Gustavo Morais, Marcus Ritter, Sergei Shudler, Torsten Hoefler, Felix Wolf: 
+Extra-P – Empirical Performance Modeling Made Easy. *Frontiers in High Performance Computing*, 3–2025, 2026. [10.3389/fhpcp.2025.1714042](https://doi.org/10.3389/fhpcp.2025.1714042)
+
+```
+@Article{calotoiu_extra-p_2026,
+  author          = {Calotoiu, Alexandru and Copik, Marcin and Czappa, Fabian and Geiß, Alexander and Morais, Gustavo and Ritter, Marcus and Shudler, Sergei and Hoefler, Torsten and Wolf, Felix},
+  title           = {Extra-{P} – {Empirical} {Performance} {Modeling} {Made} {Easy}},
+  journal         = {Frontiers in High Performance Computing},
+  year            = {2026},
+  volume          = {3–2025},
+  issn            = {2813–7337},
+  doi             = {10.3389/fhpcp.2025.1714042}
+}
+```
+
+### Selected Publications
+
+1. Alexandru Calotoiu, Torsten Hoefler, Marius Poke, Felix Wolf:
+   *Using automated performance modeling to find scalability bugs in complex codes*. In Proc. of the International
+   Conference on High Performance Computing, Networking, Storage and Analysis (SC '13), ACM, 2013.
+   [DOI](https://doi.org/10.1145/2503210.2503277)
+
+2. Alexandru Calotoiu, David Beckingsale, Christopher W. Earl, Torsten Hoefler, Ian Karlin, Martin Schulz, Felix Wolf:
+   *Fast Multi-Parameter Performance Modeling*. In Proc. of the 2016 IEEE International Conference on Cluster Computing (
    CLUSTER), Taipei, Taiwan, pages 172–181, IEEE, September
     2016. [PDF](https://apps.fz-juelich.de/jsc-pubsystem/aigaion/attachments/fastmultiparam.pdf-f839eba376c6d61a8c4cab9860b6b3bf.pdf)
 
-2. Marcus Ritter, Alexandru Calotoiu, Sebastian Rinke, Thorsten Reimann, Torsten Hoefler, Felix Wolf: *Learning
+3. Marcus Ritter, Alexandru Calotoiu, Sebastian Rinke, Thorsten Reimann, Torsten Hoefler, Felix Wolf: *Learning
    Cost-Effective Sampling Strategies for Empirical Performance Modeling.* In Proc. of the 34th IEEE International
    Parallel and Distributed Processing Symposium (IPDPS), New Orleans, LA, USA, pages 884–895, IEEE, May
     2020. [PDF](https://apps.fz-juelich.de/jsc-pubsystem/aigaion/attachments/ritter_ea_2020_ipdps.pdf-01cbe96f7a170aba7c7ef941f966d292.pdf)
 
-3. Marcus Ritter, Alexander Geiß, Johannes Wehrstein, Alexandru Calotoiu, Thorsten Reimann, Torsten Hoefler, Felix Wolf:
+4. Marcus Ritter, Alexander Geiß, Johannes Wehrstein, Alexandru Calotoiu, Thorsten Reimann, Torsten Hoefler, Felix Wolf:
    *Noise-Resilient Empirical Performance Modeling with Deep Neural Networks.* In Proc. of the 35th IEEE International
    Parallel and Distributed Processing Symposium (IPDPS), Portland, Oregon, USA, pages 23–34, IEEE, May
     2021. [PDF](http://htor.inf.ethz.ch/publications/img/noiseresilientmodeling.pdf)
+
+5. Alexander Geiß, Téodora Hovi, Alexandru Calotoiu, Felix Wolf:
+   *Validating the Performance of GPU Ports Using Differential Performance Models*. Future Generation Computer Systems, 174: 1–17, January 2026.
+   [DOI](https://doi.org/10.1016/j.future.2025.108018)
+
+7. Marcus Ritter, Benedict Naumann, Alexandru Calotoiu, Sebastian Rinke, Thorsten Reimann, Torsten Hoefler, Felix Wolf:
+   *Cost-Effective Empirical Performance Modeling*. Transactions on Parallel and Distributed Systems, 37: 575–592, February 2026.
+   [PDF](https://www.parallel.informatik.tu-darmstadt.de/zotero-lists/attachments/GH3LJG4Z_Ritter%20et%20al.%20-%202026%20-%20Cost-Effective%20Empirical%20Performance%20Modeling.pdf)
+
+9. Alexandru Calotoiu, Marcin Copik, Fabian Czappa, Alexander Geiß, Gustavo Morais, Marcus Ritter, Sergei Shudler, Torsten Hoefler, Felix Wolf:
+   *Extra-P – Empirical Performance Modeling Made Easy*. Frontiers in High Performance Computing, 3–2025, March 2026.
+   [DOI](https://doi.org/10.3389/fhpcp.2025.1714042)

{extrap-4.2.2 → extrap-4.2.4}/README.md

@@ -115,31 +115,51 @@ You can find an overview about all command line options under [docs/command-line
 
 Please cite Extra-P in your publications if it helps your research:
 
-    @inproceedings{calotoiu_ea:2013:modeling,
-      author = {Calotoiu, Alexandru and Hoefler, Torsten and Poke, Marius and Wolf, Felix},
-      month = {November},
-      title = {Using Automated Performance Modeling to Find Scalability Bugs in Complex Codes},
-      booktitle = {Proc. of the ACM/IEEE Conference on Supercomputing (SC13), Denver, CO, USA},
-      year = {2013},
-      pages = {1--12},
-      publisher = {ACM},
-      isbn = {978-1-4503-2378-9},
-      doi = {10.1145/2503210.2503277}
-    }
-
-### Publications
-
-1. Alexandru Calotoiu, David Beckingsale, Christopher W. Earl, Torsten Hoefler, Ian Karlin, Martin Schulz, Felix Wolf:
-   Fast Multi-Parameter Performance Modeling. In Proc. of the 2016 IEEE International Conference on Cluster Computing (
+Alexandru Calotoiu, Marcin Copik, Fabian Czappa, Alexander Geiß, Gustavo Morais, Marcus Ritter, Sergei Shudler, Torsten Hoefler, Felix Wolf: 
+Extra-P – Empirical Performance Modeling Made Easy. *Frontiers in High Performance Computing*, 3–2025, 2026. [10.3389/fhpcp.2025.1714042](https://doi.org/10.3389/fhpcp.2025.1714042)
+
+```
+@Article{calotoiu_extra-p_2026,
+  author          = {Calotoiu, Alexandru and Copik, Marcin and Czappa, Fabian and Geiß, Alexander and Morais, Gustavo and Ritter, Marcus and Shudler, Sergei and Hoefler, Torsten and Wolf, Felix},
+  title           = {Extra-{P} – {Empirical} {Performance} {Modeling} {Made} {Easy}},
+  journal         = {Frontiers in High Performance Computing},
+  year            = {2026},
+  volume          = {3–2025},
+  issn            = {2813–7337},
+  doi             = {10.3389/fhpcp.2025.1714042}
+}
+```
+
+### Selected Publications
+
+1. Alexandru Calotoiu, Torsten Hoefler, Marius Poke, Felix Wolf:
+   *Using automated performance modeling to find scalability bugs in complex codes*. In Proc. of the International
+   Conference on High Performance Computing, Networking, Storage and Analysis (SC '13), ACM, 2013.
+   [DOI](https://doi.org/10.1145/2503210.2503277)
+
+2. Alexandru Calotoiu, David Beckingsale, Christopher W. Earl, Torsten Hoefler, Ian Karlin, Martin Schulz, Felix Wolf:
+   *Fast Multi-Parameter Performance Modeling*. In Proc. of the 2016 IEEE International Conference on Cluster Computing (
    CLUSTER), Taipei, Taiwan, pages 172–181, IEEE, September
     2016. [PDF](https://apps.fz-juelich.de/jsc-pubsystem/aigaion/attachments/fastmultiparam.pdf-f839eba376c6d61a8c4cab9860b6b3bf.pdf)
 
-2. Marcus Ritter, Alexandru Calotoiu, Sebastian Rinke, Thorsten Reimann, Torsten Hoefler, Felix Wolf: *Learning
+3. Marcus Ritter, Alexandru Calotoiu, Sebastian Rinke, Thorsten Reimann, Torsten Hoefler, Felix Wolf: *Learning
    Cost-Effective Sampling Strategies for Empirical Performance Modeling.* In Proc. of the 34th IEEE International
    Parallel and Distributed Processing Symposium (IPDPS), New Orleans, LA, USA, pages 884–895, IEEE, May
     2020. [PDF](https://apps.fz-juelich.de/jsc-pubsystem/aigaion/attachments/ritter_ea_2020_ipdps.pdf-01cbe96f7a170aba7c7ef941f966d292.pdf)
 
-3. Marcus Ritter, Alexander Geiß, Johannes Wehrstein, Alexandru Calotoiu, Thorsten Reimann, Torsten Hoefler, Felix Wolf:
+4. Marcus Ritter, Alexander Geiß, Johannes Wehrstein, Alexandru Calotoiu, Thorsten Reimann, Torsten Hoefler, Felix Wolf:
    *Noise-Resilient Empirical Performance Modeling with Deep Neural Networks.* In Proc. of the 35th IEEE International
    Parallel and Distributed Processing Symposium (IPDPS), Portland, Oregon, USA, pages 23–34, IEEE, May
     2021. [PDF](http://htor.inf.ethz.ch/publications/img/noiseresilientmodeling.pdf)
+
+5. Alexander Geiß, Téodora Hovi, Alexandru Calotoiu, Felix Wolf:
+   *Validating the Performance of GPU Ports Using Differential Performance Models*. Future Generation Computer Systems, 174: 1–17, January 2026.
+   [DOI](https://doi.org/10.1016/j.future.2025.108018)
+
+7. Marcus Ritter, Benedict Naumann, Alexandru Calotoiu, Sebastian Rinke, Thorsten Reimann, Torsten Hoefler, Felix Wolf:
+   *Cost-Effective Empirical Performance Modeling*. Transactions on Parallel and Distributed Systems, 37: 575–592, February 2026.
+   [PDF](https://www.parallel.informatik.tu-darmstadt.de/zotero-lists/attachments/GH3LJG4Z_Ritter%20et%20al.%20-%202026%20-%20Cost-Effective%20Empirical%20Performance%20Modeling.pdf)
+
+9. Alexandru Calotoiu, Marcin Copik, Fabian Czappa, Alexander Geiß, Gustavo Morais, Marcus Ritter, Sergei Shudler, Torsten Hoefler, Felix Wolf:
+   *Extra-P – Empirical Performance Modeling Made Easy*. Frontiers in High Performance Computing, 3–2025, March 2026.
+   [DOI](https://doi.org/10.3389/fhpcp.2025.1714042)

{extrap-4.2.2 → extrap-4.2.4}/extrap/__init__.py

@@ -1,14 +1,14 @@
 # This file is part of the Extra-P software (http://www.scalasca.org/software/extra-p)
 #
-# Copyright (c) 2020-2024, Technical University of Darmstadt, Germany
+# Copyright (c) 2020-2026, Technical University of Darmstadt, Germany
 #
 # This software may be modified and distributed under the terms of a BSD-style license.
 # See the LICENSE file in the base directory for details.
 
 __title__ = "Extra-P"
-__version__ = "4.2.2"
+__version__ = "4.2.4"
 __description__ = "Extra-P, automated performance modeling for HPC applications"
-__copyright__ = "Copyright (c) 2020-2024 Technical University of Darmstadt, Darmstadt, Germany"
+__copyright__ = "Copyright (c) 2020-2026 Technical University of Darmstadt, Darmstadt, Germany"
 __documentation_link__ = f"https://github.com/extra-p/extrap/tree/v{__version__}/docs"
 __current_version_api__ = "https://pypi.org/pypi/extrap/json"
 __developed_by_html__ = 'Extra-P is developed by <a href="https://www.parallel.informatik.tu-darmstadt.de/">' \

{extrap-4.2.2 → extrap-4.2.4}/extrap/entities/functions.py

@@ -1,6 +1,6 @@
 # This file is part of the Extra-P software (http://www.scalasca.org/software/extra-p)
 #
-# Copyright (c) 2020-2024, Technical University of Darmstadt, Germany
+# Copyright (c) 2020-2026, Technical University of Darmstadt, Germany
 #
 # This software may be modified and distributed under the terms of a BSD-style license.
 # See the LICENSE file in the base directory for details.
@@ -14,7 +14,7 @@ from typing import List, Mapping, Union, Sequence
 
 import numpy
 import numpy as np
-from marshmallow import fields
+from marshmallow import fields, pre_load
 
 from extrap.entities.parameter import Parameter
 from extrap.entities.terms import CompoundTerm, MultiParameterTerm, CompoundTermSchema, MultiParameterTermSchema, \
@@ -138,7 +138,19 @@ class Function:
         elif self is other:
             return True
         else:
-            return self.__dict__ == other.__dict__
+            # Compare constant_coefficient
+            if self.constant_coefficient != other.constant_coefficient:
+                return False
+
+            # Compare compound_terms by content, not just reference
+            if len(self.compound_terms) != len(other.compound_terms):
+                return False
+
+            for t1, t2 in zip(self.compound_terms, other.compound_terms):
+                if t1 != t2:
+                    return False
+
+            return True
 
 
 class ConstantFunction(Function):
@@ -267,6 +279,23 @@ class SegmentedFunction(SingleParameterFunction):
         function_string += f" for ${parameters[0]}>={self.intervals[1][0]}$"
         return function_string
 
+    def __eq__(self, other):
+        if not isinstance(other, SegmentedFunction):
+            return NotImplemented
+        elif self is other:
+            return True
+        else:
+            # Compare segments and intervals
+            if len(self.segments) != len(other.segments):
+                return False
+            if not np.array_equal(self.intervals, other.intervals):
+                return False
+            # Compare each segment
+            for s1, s2 in zip(self.segments, other.segments):
+                if s1 != s2:
+                    return False
+            return True
+
 
 class MultiParameterFunction(Function):
     compound_terms: List[MultiParameterTerm]

{extrap-4.2.2 → extrap-4.2.4}/extrap/entities/measurement.py

@@ -1,6 +1,6 @@
 # This file is part of the Extra-P software (http://www.scalasca.org/software/extra-p)
 #
-# Copyright (c) 2020-2024, Technical University of Darmstadt, Germany
+# Copyright (c) 2020-2025, Technical University of Darmstadt, Germany
 #
 # This software may be modified and distributed under the terms of a BSD-style license.
 # See the LICENSE file in the base directory for details.
@@ -88,6 +88,8 @@ class Measurement:
             return values
         if isinstance(values, np.ndarray):
             return values
+        if all(isinstance(v, np.ndarray) for v in values):
+            return np.ma.array(values)
 
         dim_max_len = [(len(values), len(values))]
 
@@ -250,7 +252,7 @@ class MeasurementSchema(Schema):
     minimum = NumberField()
     maximum = NumberField()
     std = NumberField()
-    repetitions = fields.Int()
+    repetitions = fields.Int(allow_none=True)
     values = fields.Method('_store_values', '_load_values', allow_none=True, load_default=None)
 
     def _load_values(self, value):

{extrap-4.2.2 → extrap-4.2.4}/extrap/entities/named_entity.py

@@ -1,6 +1,6 @@
 # This file is part of the Extra-P software (http://www.scalasca.org/software/extra-p)
 #
-# Copyright (c) 2020-2022, Technical University of Darmstadt, Germany
+# Copyright (c) 2020-2025, Technical University of Darmstadt, Germany
 #
 # This software may be modified and distributed under the terms of a BSD-style license.
 # See the LICENSE file in the base directory for details.
@@ -8,7 +8,7 @@
 import copy
 import json
 from abc import abstractmethod, ABC
-from typing import Iterator, Mapping, Any
+from typing import Iterator, Any, MutableMapping
 
 from marshmallow import fields
 
@@ -65,7 +65,7 @@ class NamedEntityWithTags(NamedEntity, ABC):
 
     def __init__(self, name, **tags):
         super(NamedEntityWithTags, self).__init__(name)
-        self.tags: Mapping[str, Any] = tags or {}
+        self.tags: MutableMapping[str, Any] = tags or {}
 
     def lookup_tag(self, tag: str, default=None, prefix=1):
         if tag in self.tags:

{extrap-4.2.2 → extrap-4.2.4}/extrap/fileio/file_reader/cube_file_reader2.py

@@ -220,8 +220,9 @@ class CubeFileReader2(AbstractDirectoryReader, AbstractScalingConversionReader,
                                     total_values[callpaths[r_cnode.id]].append(cnode_values.astype(float))
                                 elif self.scaling_type == ScalingType.WEAK_THREADED:
                                     values = cnode_values.astype(float)
-                                    non_zero_value_mask = values != 0
-                                    masked_array = ma.array(values, mask=non_zero_value_mask)
+                                    if values.any():
+                                        zero_value_mask = values == 0
+                                        masked_array = ma.array(values, mask=zero_value_mask)
                                     total_values[callpaths[r_cnode.id]].append(masked_array)
                                 elif self.scaling_type == ScalingType.STRONG:
                                     total_values[callpaths[r_cnode.id]].append(cnode_values.sum().astype(float))
@@ -241,8 +242,11 @@ class CubeFileReader2(AbstractDirectoryReader, AbstractScalingConversionReader,
                                 aggregated_values[(callpath, metric)].append(cnode_values.astype(float))
                             elif self.scaling_type == ScalingType.WEAK_THREADED:
                                 values = cnode_values.astype(float)
-                                non_zero_value_mask = values != 0
-                                aggregated_values[(callpath, metric)].append(ma.array(values, mask=non_zero_value_mask))
+                                if not values.any():
+                                    aggregated_values[(callpath, metric)].append(ma.array(values))
+                                else:
+                                    zero_value_mask = (values == 0)
+                                    aggregated_values[(callpath, metric)].append(ma.array(values, mask=zero_value_mask))
                             # in case of strong scaling calculate the sum over all mpi process values
                             elif self.scaling_type == ScalingType.STRONG:
                                 aggregated_values[(callpath, metric)].append(cnode_values.sum().astype(float))

{extrap-4.2.2 → extrap-4.2.4}/extrap/gui/GraphWidget.py

@@ -1,6 +1,6 @@
 # This file is part of the Extra-P software (http://www.scalasca.org/software/extra-p)
 #
-# Copyright (c) 2020-2024, Technical University of Darmstadt, Germany
+# Copyright (c) 2020-2025, Technical University of Darmstadt, Germany
 #
 # This software may be modified and distributed under the terms of a BSD-style license.
 # See the LICENSE file in the base directory for details.
@@ -16,7 +16,6 @@ import numpy
 from PySide6.QtCore import *  # @UnusedWildImport
 from PySide6.QtGui import *  # @UnusedWildImport
 from PySide6.QtWidgets import *  # @UnusedWildImport
-
 from extrap.gui.Utils import formatFormula
 from extrap.gui.Utils import formatNumber
 from extrap.gui.plots.AbstractPlotWidget import AbstractPlotWidget
@@ -569,9 +568,8 @@ class GraphWidget(QWidget):
             number_of_x_points, x_list, x_values = self._calculate_evaluation_points(length_x_axis, x_min, x_max)
         else:
             number_of_x_points, x_list, x_values = self._calculate_evaluation_points(length_x_axis)
-        previous = numpy.seterr(invalid='ignore', divide='ignore')
-        y_list = function.evaluate(x_list).reshape(-1)
-        numpy.seterr(**previous)
+        with numpy.errstate(invalid='ignore', divide='ignore'):
+            y_list = function.evaluate(x_list).reshape(-1)
         cord_list = self._create_drawing_iterator(x_values, y_list)
 
         return cord_list
@@ -585,10 +583,9 @@ class GraphWidget(QWidget):
 
         y_list = numpy.zeros(number_of_x_points)
 
-        previous = numpy.seterr(invalid='ignore', divide='ignore')
-        for function in functions:
-            y_list += function.evaluate(x_list).reshape(-1)
-        numpy.seterr(**previous)
+        with numpy.errstate(invalid='ignore', divide='ignore'):
+            for function in functions:
+                y_list += function.evaluate(x_list).reshape(-1)
 
         cord_list = self._create_drawing_iterator(x_values, y_list)
 
@@ -741,43 +738,41 @@ class GraphWidget(QWidget):
                 y = max(model.predictions)
                 y_max = max(y, y_max)
 
-        previous = numpy.seterr(invalid='ignore', divide='ignore')
-
-        if self.combine_all_callpath:
-            y_agg = 0
+        with numpy.errstate(invalid='ignore', divide='ignore'):
+            if self.combine_all_callpath:
+                y_agg = 0
+                for model in modelList:
+                    function = model.hypothesis.function
+                    y_agg = y_agg + function.evaluate(pv_list)
+                y_max = max(y_agg, y_max)
+
+                pv_list[param] = 1
+                y_agg = 0
+                for model in modelList:
+                    function = model.hypothesis.function
+                    y = function.evaluate(pv_list)
+                    if math.isinf(y):
+                        y = max(model.predictions)
+                    y_agg += y
+                y_max = max(y_agg, y_max)
+
+            # Check the value at the end of the displayed interval
             for model in modelList:
                 function = model.hypothesis.function
-                y_agg = y_agg + function.evaluate(pv_list)
-            y_max = max(y_agg, y_max)
+                y = function.evaluate(pv_list)
+                if math.isinf(y):
+                    y = max(model.predictions)
+                y_max = max(y, y_max)
 
+            # Check the value at the beginning of the displayed interval
             pv_list[param] = 1
-            y_agg = 0
             for model in modelList:
                 function = model.hypothesis.function
                 y = function.evaluate(pv_list)
                 if math.isinf(y):
                     y = max(model.predictions)
-                y_agg += y
-            y_max = max(y_agg, y_max)
-
-        # Check the value at the end of the displayed interval
-        for model in modelList:
-            function = model.hypothesis.function
-            y = function.evaluate(pv_list)
-            if math.isinf(y):
-                y = max(model.predictions)
-            y_max = max(y, y_max)
-
-        # Check the value at the beginning of the displayed interval
-        pv_list[param] = 1
-        for model in modelList:
-            function = model.hypothesis.function
-            y = function.evaluate(pv_list)
-            if math.isinf(y):
-                y = max(model.predictions)
-            y_max = max(y, y_max)
+                y_max = max(y, y_max)
 
-        numpy.seterr(**previous)
         # Ensure that the maximum value is never too small
         if y_max < 0.000001:
             y_max = 1

{extrap-4.2.2 → extrap-4.2.4}/extrap/gui/SelectorWidget.py

@@ -1,6 +1,6 @@
 # This file is part of the Extra-P software (http://www.scalasca.org/software/extra-p)
 #
-# Copyright (c) 2020-2024, Technical University of Darmstadt, Germany
+# Copyright (c) 2020-2025, Technical University of Darmstadt, Germany
 #
 # This software may be modified and distributed under the terms of a BSD-style license.
 # See the LICENSE file in the base directory for details.
@@ -11,9 +11,7 @@ import math
 from typing import Optional, Sequence, TYPE_CHECKING, Tuple
 
 import numpy
-from PySide6.QtCore import Slot
 from PySide6.QtWidgets import *  # @UnusedWildImport
-
 from extrap.entities.calltree import Node
 from extrap.entities.metric import Metric
 from extrap.entities.model import Model
@@ -346,9 +344,9 @@ class SelectorWidget(QWidget):
                 param_value_list = self.getParameterValues()
                 call_tree = experiment.call_tree
                 nodes = call_tree.get_nodes()
-                previous = numpy.seterr(divide='ignore', invalid='ignore')
-                value_list = self.iterate_children(model_set.models, param_value_list, nodes, selected_metric)
-                numpy.seterr(**previous)
+                with numpy.errstate(divide='ignore', invalid='ignore'):
+                    value_list = self.iterate_children(model_set.models, param_value_list, nodes, selected_metric)
+
                 if len(value_list) > 0:
                     min_max_value = max(0.0, min(value_list)), max(0.0, max(value_list))
         self.min_value, self.max_value = min_max_value

{extrap-4.2.2 → extrap-4.2.4}/extrap/gui/TreeModel.py

@@ -1,6 +1,6 @@
 # This file is part of the Extra-P software (http://www.scalasca.org/software/extra-p)
 #
-# Copyright (c) 2020-2024, Technical University of Darmstadt, Germany
+# Copyright (c) 2020-2025, Technical University of Darmstadt, Germany
 #
 # This software may be modified and distributed under the terms of a BSD-style license.
 # See the LICENSE file in the base directory for details.
@@ -8,11 +8,11 @@
 from __future__ import annotations
 
 import copy
-import numpy
-from PySide6.QtCore import *  # @UnusedWildImport
 from enum import Enum, auto
 from typing import Optional, TYPE_CHECKING, List, Callable, Any
 
+import numpy
+from PySide6.QtCore import *  # @UnusedWildImport
 from extrap.entities import calltree
 from extrap.entities.calltree import CallTree, Node
 from extrap.entities.experiment import Experiment
@@ -148,12 +148,13 @@ class TreeModel(QAbstractItemModel):
                 return formatFormula(formula.to_string(*parameters))
             else:
                 parameters = self.selector_widget.getParameterValues()
-                previous = numpy.seterr(divide='ignore', invalid='ignore')
-                if role == Qt.ToolTipRole and isinstance(model, SegmentedModel):
-                    res = _format_number_segmented_model(model, lambda m: m.hypothesis.function.evaluate(parameters))
-                else:
-                    res = formatNumber(str(formula.evaluate(parameters)))
-                numpy.seterr(**previous)
+                with numpy.errstate(divide='ignore', invalid='ignore'):
+                    if role == Qt.ToolTipRole and isinstance(model, SegmentedModel):
+                        res = _format_number_segmented_model(model,
+                                                             lambda m: m.hypothesis.function.evaluate(parameters))
+                    else:
+                        res = formatNumber(str(formula.evaluate(parameters)))
+
                 return res
         elif index.column() == 4:
             if role == Qt.ToolTipRole and isinstance(model, SegmentedModel):
@@ -180,9 +181,8 @@ class TreeModel(QAbstractItemModel):
     def get_comparison_value(self, model):
         parameters = self.selector_widget.getParameterValues()
         formula = model.hypothesis.function
-        previous = numpy.seterr(divide='ignore', invalid='ignore')
-        value = formula.evaluate(parameters)
-        numpy.seterr(**previous)
+        with numpy.errstate(divide='ignore', invalid='ignore'):
+            value = formula.evaluate(parameters)
         return value
 
     def getSelectedModel(self, callpath) -> tuple[Optional[Model], Experiment]:

{extrap-4.2.2 → extrap-4.2.4}/extrap/gui/components/ExpanderWidget.py

@@ -1,6 +1,6 @@
 # This file is part of the Extra-P software (http://www.scalasca.org/software/extra-p)
 #
-# Copyright (c) 2020-2024, Technical University of Darmstadt, Germany
+# Copyright (c) 2020-2025, Technical University of Darmstadt, Germany
 #
 # This software may be modified and distributed under the terms of a BSD-style license.
 # See the LICENSE file in the base directory for details.
@@ -44,7 +44,7 @@ class ExpanderWidget(QGroupBox):
 
     def initUI(self):
         self.setStyleSheet('ExpanderWidget{ '
-                           '    padding:2px -1px -1px 0px; margin-left:-1px; '
+                           '    padding:0px -1px -1px 0px; margin-left:-1px; '
                            f'{" margin-top: -19px; padding-top:19px; " if sys.platform.startswith("darwin") else ""}'
                            '}'
                            'ExpanderWidget::title{ '
@@ -97,7 +97,6 @@ class ExpanderWidget(QGroupBox):
         self._content.setSizePolicy(QSizePolicy.Expanding, QSizePolicy.Fixed)
         self._content.setMinimumHeight(0)
         self._content.setMaximumHeight(0)
-        self._content
 
         layout.addWidget(self._toggle)
         layout.addWidget(self._content)

{extrap-4.2.2 → extrap-4.2.4}/extrap/gui/plots/BaseGraphWidget.py

@@ -1,6 +1,6 @@
 # This file is part of the Extra-P software (http://www.scalasca.org/software/extra-p)
 #
-# Copyright (c) 2020-2024, Technical University of Darmstadt, Germany
+# Copyright (c) 2020-2025, Technical University of Darmstadt, Germany
 #
 # This software may be modified and distributed under the terms of a BSD-style license.
 # See the LICENSE file in the base directory for details.
@@ -13,13 +13,12 @@ from typing import TYPE_CHECKING
 import matplotlib
 import numpy as np
 from PySide6.QtWidgets import QSizePolicy
+from extrap.util.formatting_helper import replace_method_parameters
 from matplotlib import patches as mpatches
 from matplotlib.backends.backend_qt5agg import FigureCanvasQTAgg as FigureCanvas
 from matplotlib.figure import Figure
 from mpl_toolkits.mplot3d import Axes3D
 
-from extrap.util.formatting_helper import replace_method_parameters
-
 if TYPE_CHECKING:
     from extrap.gui.MainWidget import MainWidget
 
@@ -105,15 +104,14 @@ class GraphDisplayWindow(FigureCanvas):
         # Get the z value for the x and y value
         z_List = list()
         Z_List = list()
-        previous = np.seterr(invalid='ignore', divide='ignore')
-        for model in model_list:
-            function = model.hypothesis.function
-            zs = self.calculate_z_optimized(X, Y, function)
-            Z = zs.reshape(X.shape)
-            z_List.append(zs)
-            Z_List.append(Z)
-            max_z = max(max_z, np.max(zs[np.logical_not(np.isinf(zs))]))
-        np.seterr(**previous)
+        with np.errstate(invalid='ignore', divide='ignore'):
+            for model in model_list:
+                function = model.hypothesis.function
+                zs = self.calculate_z_optimized(X, Y, function)
+                Z = zs.reshape(X.shape)
+                z_List.append(zs)
+                Z_List.append(Z)
+                max_z = max(max_z, np.max(zs[np.logical_not(np.isinf(zs))]))
         for z, Z in zip(z_List, Z_List):
             z[np.isinf(z)] = max_z
             Z[np.isinf(Z)] = max_z