sdgym 0.9.1.dev0__tar.gz → 0.10.0.dev0__tar.gz

{sdgym-0.9.1.dev0 → sdgym-0.10.0.dev0}/PKG-INFO

@@ -1,6 +1,6 @@
-Metadata-Version: 2.1
+Metadata-Version: 2.2
 Name: sdgym
-Version: 0.9.1.dev0
+Version: 0.10.0.dev0
 Summary: Benchmark tabular synthetic data generators using a variety of datasets
 Author-email: "DataCebo, Inc." <info@sdv.dev>
 License: BSL-1.1
@@ -30,9 +30,9 @@ Requires-Dist: botocore<2,>=1.31
 Requires-Dist: cloudpickle>=2.1.0
 Requires-Dist: compress-pickle>=1.2.0
 Requires-Dist: humanfriendly>=8.2
-Requires-Dist: numpy<2.0.0,>=1.21.0; python_version < "3.10"
-Requires-Dist: numpy<2.0.0,>=1.23.3; python_version >= "3.10" and python_version < "3.12"
-Requires-Dist: numpy<2.0.0,>=1.26.0; python_version >= "3.12"
+Requires-Dist: numpy>=1.21.6; python_version < "3.10"
+Requires-Dist: numpy>=1.23.3; python_version >= "3.10" and python_version < "3.12"
+Requires-Dist: numpy>=1.26.0; python_version >= "3.12"
 Requires-Dist: pandas>=1.4.0; python_version < "3.11"
 Requires-Dist: pandas>=1.5.0; python_version >= "3.11" and python_version < "3.12"
 Requires-Dist: pandas>=2.1.1; python_version >= "3.12"
@@ -45,18 +45,23 @@ Requires-Dist: scipy>=1.7.3; python_version < "3.10"
 Requires-Dist: scipy>=1.9.2; python_version >= "3.10" and python_version < "3.12"
 Requires-Dist: scipy>=1.12.0; python_version >= "3.12"
 Requires-Dist: tabulate<0.9,>=0.8.3
-Requires-Dist: torch>=1.9.0; python_version < "3.10"
+Requires-Dist: torch>=1.12.1; python_version < "3.10"
 Requires-Dist: torch>=2.0.0; python_version >= "3.10" and python_version < "3.12"
 Requires-Dist: torch>=2.2.0; python_version >= "3.12"
-Requires-Dist: tqdm>=4.29
+Requires-Dist: tqdm>=4.66.3
 Requires-Dist: XlsxWriter>=1.2.8
-Requires-Dist: rdt>=1.12.1
-Requires-Dist: sdmetrics>=0.14.1
-Requires-Dist: sdv>=1.13.1
+Requires-Dist: rdt>=1.13.1
+Requires-Dist: sdmetrics>=0.17.0
+Requires-Dist: sdv>=1.17.2
 Provides-Extra: dask
 Requires-Dist: dask; extra == "dask"
 Requires-Dist: distributed; extra == "dask"
+Provides-Extra: realtabformer
+Requires-Dist: realtabformer>=0.2.2; extra == "realtabformer"
+Requires-Dist: torch>=2.0.0; (python_version >= "3.8" and python_version < "3.12") and extra == "realtabformer"
+Requires-Dist: torch>=2.2.0; python_version >= "3.12" and extra == "realtabformer"
 Provides-Extra: test
+Requires-Dist: sdgym[realtabformer]; extra == "test"
 Requires-Dist: pytest>=6.2.5; extra == "test"
 Requires-Dist: pytest-cov>=2.6.0; extra == "test"
 Requires-Dist: jupyter<2,>=1.0.0; extra == "test"

{sdgym-0.9.1.dev0 → sdgym-0.10.0.dev0}/pyproject.toml

@@ -27,9 +27,9 @@ dependencies = [
     'cloudpickle>=2.1.0',
     'compress-pickle>=1.2.0',
     'humanfriendly>=8.2',
-    "numpy>=1.21.0,<2.0.0;python_version<'3.10'",
-    "numpy>=1.23.3,<2.0.0;python_version>='3.10' and python_version<'3.12'",
-    "numpy>=1.26.0,<2.0.0;python_version>='3.12'",
+    "numpy>=1.21.6;python_version<'3.10'",
+    "numpy>=1.23.3;python_version>='3.10' and python_version<'3.12'",
+    "numpy>=1.26.0;python_version>='3.12'",
     "pandas>=1.4.0;python_version<'3.11'",
     "pandas>=1.5.0;python_version>='3.11' and python_version<'3.12'",
     "pandas>=2.1.1;python_version>='3.12'",
@@ -42,14 +42,14 @@ dependencies = [
     "scipy>=1.9.2;python_version>='3.10' and python_version<'3.12'",
     "scipy>=1.12.0;python_version>='3.12'",
     'tabulate>=0.8.3,<0.9',
-    "torch>=1.9.0;python_version<'3.10'",
+    "torch>=1.12.1;python_version<'3.10'",
     "torch>=2.0.0;python_version>='3.10' and python_version<'3.12'",
     "torch>=2.2.0;python_version>='3.12'",
-    'tqdm>=4.29',
+    'tqdm>=4.66.3',
     'XlsxWriter>=1.2.8',
-    'rdt>=1.12.1',
-    'sdmetrics>=0.14.1',
-    'sdv>=1.13.1',
+    'rdt>=1.13.1',
+    'sdmetrics>=0.17.0',
+    'sdv>=1.17.2',
 ]
 
 [project.urls]
@@ -64,7 +64,13 @@ sdgym = { main = 'sdgym.cli.__main__:main' }
 
 [project.optional-dependencies]
 dask = ['dask', 'distributed']
+realtabformer = [
+    'realtabformer>=0.2.2',
+    "torch>=2.0.0;python_version>='3.8' and python_version<'3.12'",
+    "torch>=2.2.0;python_version>='3.12'",
+]
 test = [
+    'sdgym[realtabformer]',
     'pytest>=6.2.5',
     'pytest-cov>=2.6.0',
     'jupyter>=1.0.0,<2',
@@ -134,7 +140,7 @@ namespaces = false
 version = {attr = 'sdgym.__version__'}
 
 [tool.bumpversion]
-current_version = "0.9.1.dev0"
+current_version = "0.10.0.dev0"
 parse = '(?P<major>\d+)\.(?P<minor>\d+)\.(?P<patch>\d+)(\.(?P<release>[a-z]+)(?P<candidate>\d+))?'
 serialize = [
     '{major}.{minor}.{patch}.{release}{candidate}',
@@ -198,10 +204,11 @@ select = [
     # print statements
     "T201",
     # pandas-vet
-    "PD"
+    "PD",
+    # numpy 2.0
+    "NPY201"
 ]
 ignore = [
-    "E501",
     # pydocstyle
     "D107",  # Missing docstring in __init__
     "D417",   # Missing argument descriptions in the docstring, this is a bug from pydocstyle: https://github.com/PyCQA/pydocstyle/issues/449
@@ -230,4 +237,4 @@ convention = "google"
 
 [tool.ruff.lint.pycodestyle]
 max-doc-length = 100
-max-line-length = 100
+max-line-length = 100

{sdgym-0.9.1.dev0 → sdgym-0.10.0.dev0}/sdgym/__init__.py

@@ -8,7 +8,7 @@ __author__ = 'DataCebo, Inc.'
 __copyright__ = 'Copyright (c) 2022 DataCebo, Inc.'
 __email__ = 'info@sdv.dev'
 __license__ = 'BSL-1.1'
-__version__ = '0.9.1.dev0'
+__version__ = '0.10.0.dev0'
 
 import logging
 

{sdgym-0.9.1.dev0 → sdgym-0.10.0.dev0}/sdgym/benchmark.py

@@ -66,8 +66,7 @@ N_BYTES_IN_MB = 1000 * 1000
 def _validate_inputs(output_filepath, detailed_results_folder, synthesizers, custom_synthesizers):
     if output_filepath and os.path.exists(output_filepath):
         raise ValueError(
-            f'{output_filepath} already exists. '
-            'Please provide a file that does not already exist.'
+            f'{output_filepath} already exists. Please provide a file that does not already exist.'
         )
 
     if detailed_results_folder and os.path.exists(detailed_results_folder):
@@ -149,14 +148,14 @@ def _generate_job_args_list(
 def _synthesize(synthesizer_dict, real_data, metadata):
     synthesizer = synthesizer_dict['synthesizer']
     if isinstance(synthesizer, type):
-        assert issubclass(
-            synthesizer, BaselineSynthesizer
-        ), '`synthesizer` must be a synthesizer class'
+        assert issubclass(synthesizer, BaselineSynthesizer), (
+            '`synthesizer` must be a synthesizer class'
+        )
         synthesizer = synthesizer()
     else:
-        assert issubclass(
-            type(synthesizer), BaselineSynthesizer
-        ), '`synthesizer` must be an instance of a synthesizer class.'
+        assert issubclass(type(synthesizer), BaselineSynthesizer), (
+            '`synthesizer` must be an instance of a synthesizer class.'
+        )
 
     get_synthesizer = synthesizer.get_trained_synthesizer
     sample_from_synthesizer = synthesizer.sample_from_synthesizer
@@ -747,6 +746,7 @@ def benchmark_single_table(
                 - ``CTGANSynthesizer``
                 - ``CopulaGANSynthesizer``
                 - ``TVAESynthesizer``
+                - ``RealTabFormerSynthesizer``
 
         custom_synthesizers (list[class] or ``None``):
             A list of custom synthesizer classes to use. These can be completely custom or

{sdgym-0.9.1.dev0 → sdgym-0.10.0.dev0}/sdgym/cli/__main__.py

@@ -180,9 +180,9 @@ def _get_parser():
     )
     run.add_argument('-m', '--metrics', nargs='+', help='Metrics to apply. Accepts multiple names.')
     run.add_argument('-b', '--bucket', help='Bucket from which to download the datasets.')
-    run.add_argument('-dp' '--datasets-path', help='Path where datasets can be found.')
+    run.add_argument('-dp--datasets-path', help='Path where datasets can be found.')
     run.add_argument(
-        '-dm' '--modalities', nargs='+', help='Data Modalities to run. Accepts multiple names.'
+        '-dm--modalities', nargs='+', help='Data Modalities to run. Accepts multiple names.'
     )
     run.add_argument('-i', '--iterations', type=int, default=1, help='Number of iterations.')
     run.add_argument(
@@ -219,13 +219,13 @@ def _get_parser():
         '-g', '--groupby', nargs='+', help='Group scores leaderboard by the given fields.'
     )
     run.add_argument(
-        '-ak' '--aws-key',
+        '-ak--aws-key',
         type=str,
         required=False,
         help='Aws access key ID to use when reading datasets.',
     )
     run.add_argument(
-        '-as' '--aws-secret',
+        '-as--aws-secret',
         type=str,
         required=False,
         help='Aws secret access key to use when reading datasets.',
@@ -234,10 +234,10 @@ def _get_parser():
         '-j', '--jobs', type=str, required=False, help='Serialized list of jobs to run.'
     )
     run.add_argument(
-        '-mr' '--max-rows', type=int, help='Cap the number of rows to model from each dataset.'
+        '-mr--max-rows', type=int, help='Cap the number of rows to model from each dataset.'
     )
     run.add_argument(
-        '-mc' '--max-columns',
+        '-mc--max-columns',
         type=int,
         help='Cap the number of columns to model from each dataset.',
     )

{sdgym-0.9.1.dev0 → sdgym-0.10.0.dev0}/sdgym/synthesizers/__init__.py

@@ -9,6 +9,7 @@ from sdgym.synthesizers.generate import (
 )
 from sdgym.synthesizers.identity import DataIdentity
 from sdgym.synthesizers.column import ColumnSynthesizer
+from sdgym.synthesizers.realtabformer import RealTabFormerSynthesizer
 from sdgym.synthesizers.sdv import (
     CopulaGANSynthesizer,
     CTGANSynthesizer,
@@ -38,4 +39,5 @@ __all__ = (
     'create_sdv_synthesizer_variant',
     'create_sequential_synthesizer',
     'SYNTHESIZER_MAPPING',
+    'RealTabFormerSynthesizer',
 )

{sdgym-0.9.1.dev0 → sdgym-0.10.0.dev0}/sdgym/synthesizers/base.py

@@ -2,9 +2,9 @@
 
 import abc
 import logging
+import warnings
 
-from sdv.metadata.multi_table import MultiTableMetadata
-from sdv.metadata.single_table import SingleTableMetadata
+from sdv.metadata import Metadata
 
 LOGGER = logging.getLogger(__name__)
 
@@ -54,8 +54,11 @@ class BaselineSynthesizer(abc.ABC):
             obj:
                 The synthesizer object.
         """
-        metadata_class = MultiTableMetadata() if 'tables' in metadata else SingleTableMetadata()
-        metadata = metadata_class.load_from_dict(metadata)
+        metadata_object = Metadata()
+        with warnings.catch_warnings():
+            warnings.simplefilter('ignore', UserWarning)
+            metadata = metadata_object.load_from_dict(metadata)
+
         return self._get_trained_synthesizer(data, metadata)
 
     def sample_from_synthesizer(self, synthesizer, n_samples):

{sdgym-0.9.1.dev0 → sdgym-0.10.0.dev0}/sdgym/synthesizers/column.py

@@ -1,11 +1,16 @@
 """ColumnSynthesizer module."""
 
+import logging
+
 import pandas as pd
 from rdt.hyper_transformer import HyperTransformer
+from sdv.metadata import Metadata
 from sklearn.mixture import GaussianMixture
 
 from sdgym.synthesizers.base import BaselineSynthesizer
 
+LOGGER = logging.getLogger(__name__)
+
 
 class ColumnSynthesizer(BaselineSynthesizer):
     """Synthesizer that learns each column independently.
@@ -17,6 +22,27 @@ class ColumnSynthesizer(BaselineSynthesizer):
     def _get_trained_synthesizer(self, real_data, metadata):
         hyper_transformer = HyperTransformer()
         hyper_transformer.detect_initial_config(real_data)
+        supported_sdtypes = hyper_transformer._get_supported_sdtypes()
+        config = {}
+        if isinstance(metadata, Metadata):
+            table_name = metadata._get_single_table_name()
+            columns = metadata.tables[table_name].columns
+        else:
+            columns = metadata.columns
+
+        for column_name, column in columns.items():
+            sdtype = column['sdtype']
+            if sdtype in supported_sdtypes:
+                config[column_name] = sdtype
+            elif column.get('pii', False):
+                config[column_name] = 'pii'
+            else:
+                LOGGER.info(
+                    f'Column {column} sdtype: {sdtype} is not supported, '
+                    f'defaulting to inferred type.'
+                )
+
+        hyper_transformer.update_sdtypes(config)
 
         # This is done to match the behavior of the synthesizer for SDGym <= 0.6.0
         columns_to_remove = [

{sdgym-0.9.1.dev0 → sdgym-0.10.0.dev0}/sdgym/synthesizers/generate.py

@@ -49,7 +49,7 @@ def create_sdv_synthesizer_variant(display_name, synthesizer_class, synthesizer_
     if synthesizer_class not in SYNTHESIZER_MAPPING.keys():
         raise ValueError(
             f'Synthesizer class {synthesizer_class} is not recognized. '
-            f"The supported options are {', '.join(SYNTHESIZER_MAPPING.keys())}"
+            f'The supported options are {", ".join(SYNTHESIZER_MAPPING.keys())}'
         )
 
     baseclass = SDVTabularSynthesizer

sdgym-0.10.0.dev0/sdgym/synthesizers/realtabformer.py

@@ -0,0 +1,46 @@
+"""REaLTabFormer integration."""
+
+import contextlib
+import logging
+from functools import partialmethod
+
+import tqdm
+
+from sdgym.synthesizers.base import BaselineSynthesizer
+
+
+@contextlib.contextmanager
+def prevent_tqdm_output():
+    """Temporarily disables tqdm m."""
+    tqdm.__init__ = partialmethod(tqdm.__init__, disable=True)
+    try:
+        yield
+    finally:
+        tqdm.__init__ = partialmethod(tqdm.__init__, disable=False)
+
+
+class RealTabFormerSynthesizer(BaselineSynthesizer):
+    """Custom wrapper for the REaLTabFormer synthesizer to make it work with SDGym."""
+
+    LOGGER = logging.getLogger(__name__)
+    _MODEL_KWARGS = None
+
+    def _get_trained_synthesizer(self, data, metadata):
+        try:
+            from realtabformer import REaLTabFormer
+        except Exception as exception:
+            raise ValueError(
+                "In order to use 'RealTabFormerSynthesizer' you have to install the extra"
+                " dependencies by running  pip install sdgym['realtabformer'] "
+            ) from exception
+
+        with prevent_tqdm_output():
+            model_kwargs = self._MODEL_KWARGS.copy() if self._MODEL_KWARGS else {}
+            model = REaLTabFormer(model_type='tabular', **model_kwargs)
+            model.fit(data, device='cpu')
+
+        return model
+
+    def _sample_from_synthesizer(self, synthesizer, n_sample):
+        """Sample synthetic data with specified sample count."""
+        return synthesizer.sample(n_sample, device='cpu')

{sdgym-0.9.1.dev0 → sdgym-0.10.0.dev0}/sdgym/synthesizers/uniform.py

@@ -19,7 +19,8 @@ class UniformSynthesizer(BaselineSynthesizer):
         hyper_transformer.detect_initial_config(real_data)
         supported_sdtypes = hyper_transformer._get_supported_sdtypes()
         config = {}
-        for column_name, column in metadata.columns.items():
+        table = next(iter(metadata.tables.values()))
+        for column_name, column in table.columns.items():
             sdtype = column['sdtype']
             if sdtype in supported_sdtypes:
                 config[column_name] = sdtype
@@ -27,7 +28,7 @@ class UniformSynthesizer(BaselineSynthesizer):
                 config[column_name] = 'pii'
             else:
                 LOGGER.info(
-                    f'Column {column} sdtype: {sdtype} is not supported, '
+                    f'Column {column_name} sdtype: {sdtype} is not supported, '
                     f'defaulting to inferred type.'
                 )
 

{sdgym-0.9.1.dev0 → sdgym-0.10.0.dev0}/sdgym.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
-Metadata-Version: 2.1
+Metadata-Version: 2.2
 Name: sdgym
-Version: 0.9.1.dev0
+Version: 0.10.0.dev0
 Summary: Benchmark tabular synthetic data generators using a variety of datasets
 Author-email: "DataCebo, Inc." <info@sdv.dev>
 License: BSL-1.1
@@ -30,9 +30,9 @@ Requires-Dist: botocore<2,>=1.31
 Requires-Dist: cloudpickle>=2.1.0
 Requires-Dist: compress-pickle>=1.2.0
 Requires-Dist: humanfriendly>=8.2
-Requires-Dist: numpy<2.0.0,>=1.21.0; python_version < "3.10"
-Requires-Dist: numpy<2.0.0,>=1.23.3; python_version >= "3.10" and python_version < "3.12"
-Requires-Dist: numpy<2.0.0,>=1.26.0; python_version >= "3.12"
+Requires-Dist: numpy>=1.21.6; python_version < "3.10"
+Requires-Dist: numpy>=1.23.3; python_version >= "3.10" and python_version < "3.12"
+Requires-Dist: numpy>=1.26.0; python_version >= "3.12"
 Requires-Dist: pandas>=1.4.0; python_version < "3.11"
 Requires-Dist: pandas>=1.5.0; python_version >= "3.11" and python_version < "3.12"
 Requires-Dist: pandas>=2.1.1; python_version >= "3.12"
@@ -45,18 +45,23 @@ Requires-Dist: scipy>=1.7.3; python_version < "3.10"
 Requires-Dist: scipy>=1.9.2; python_version >= "3.10" and python_version < "3.12"
 Requires-Dist: scipy>=1.12.0; python_version >= "3.12"
 Requires-Dist: tabulate<0.9,>=0.8.3
-Requires-Dist: torch>=1.9.0; python_version < "3.10"
+Requires-Dist: torch>=1.12.1; python_version < "3.10"
 Requires-Dist: torch>=2.0.0; python_version >= "3.10" and python_version < "3.12"
 Requires-Dist: torch>=2.2.0; python_version >= "3.12"
-Requires-Dist: tqdm>=4.29
+Requires-Dist: tqdm>=4.66.3
 Requires-Dist: XlsxWriter>=1.2.8
-Requires-Dist: rdt>=1.12.1
-Requires-Dist: sdmetrics>=0.14.1
-Requires-Dist: sdv>=1.13.1
+Requires-Dist: rdt>=1.13.1
+Requires-Dist: sdmetrics>=0.17.0
+Requires-Dist: sdv>=1.17.2
 Provides-Extra: dask
 Requires-Dist: dask; extra == "dask"
 Requires-Dist: distributed; extra == "dask"
+Provides-Extra: realtabformer
+Requires-Dist: realtabformer>=0.2.2; extra == "realtabformer"
+Requires-Dist: torch>=2.0.0; (python_version >= "3.8" and python_version < "3.12") and extra == "realtabformer"
+Requires-Dist: torch>=2.2.0; python_version >= "3.12" and extra == "realtabformer"
 Provides-Extra: test
+Requires-Dist: sdgym[realtabformer]; extra == "test"
 Requires-Dist: pytest>=6.2.5; extra == "test"
 Requires-Dist: pytest-cov>=2.6.0; extra == "test"
 Requires-Dist: jupyter<2,>=1.0.0; extra == "test"

{sdgym-0.9.1.dev0 → sdgym-0.10.0.dev0}/sdgym.egg-info/SOURCES.txt

@@ -25,6 +25,7 @@ sdgym/synthesizers/base.py
 sdgym/synthesizers/column.py
 sdgym/synthesizers/generate.py
 sdgym/synthesizers/identity.py
+sdgym/synthesizers/realtabformer.py
 sdgym/synthesizers/sdv.py
 sdgym/synthesizers/uniform.py
 tests/test_tasks.py

{sdgym-0.9.1.dev0 → sdgym-0.10.0.dev0}/sdgym.egg-info/requires.txt

@@ -6,17 +6,17 @@ compress-pickle>=1.2.0
 humanfriendly>=8.2
 psutil>=5.7
 tabulate<0.9,>=0.8.3
-tqdm>=4.29
+tqdm>=4.66.3
 XlsxWriter>=1.2.8
-rdt>=1.12.1
-sdmetrics>=0.14.1
-sdv>=1.13.1
+rdt>=1.13.1
+sdmetrics>=0.17.0
+sdv>=1.17.2
 
 [:python_version < "3.10"]
-numpy<2.0.0,>=1.21.0
+numpy>=1.21.6
 scikit-learn>=1.0.2
 scipy>=1.7.3
-torch>=1.9.0
+torch>=1.12.1
 
 [:python_version < "3.11"]
 pandas>=1.4.0
@@ -25,7 +25,7 @@ pandas>=1.4.0
 scikit-learn>=1.1.0
 
 [:python_version >= "3.10" and python_version < "3.12"]
-numpy<2.0.0,>=1.23.3
+numpy>=1.23.3
 scipy>=1.9.2
 torch>=2.0.0
 
@@ -34,7 +34,7 @@ pandas>=1.5.0
 scikit-learn>=1.1.3
 
 [:python_version >= "3.12"]
-numpy<2.0.0,>=1.26.0
+numpy>=1.26.0
 pandas>=2.1.1
 scikit-learn>=1.3.1
 scipy>=1.12.0
@@ -61,7 +61,17 @@ tox<5,>=2.9.1
 importlib-metadata>=3.6
 invoke
 
+[realtabformer]
+realtabformer>=0.2.2
+
+[realtabformer:python_version >= "3.12"]
+torch>=2.2.0
+
+[realtabformer:python_version >= "3.8" and python_version < "3.12"]
+torch>=2.0.0
+
 [test]
+sdgym[realtabformer]
 pytest>=6.2.5
 pytest-cov>=2.6.0
 jupyter<2,>=1.0.0

sdgym-0.10.0.dev0/tests/test_tasks.py

@@ -0,0 +1,207 @@
+"""Tests for the ``tasks.py`` file."""
+
+from tasks import _get_extra_dependencies, _get_minimum_versions, _resolve_version_conflicts
+
+
+def test_get_minimum_versions():
+    """Test the ``_get_minimum_versions`` method.
+
+    The method should return the minimum versions of the dependencies for the given python version.
+    If a library is linked to an URL, the minimum version should be the URL.
+    """
+    # Setup
+    dependencies = [
+        "numpy>=1.20.0,<2;python_version<'3.10'",
+        "numpy>=1.23.3,<2;python_version>='3.10'",
+        "pandas>=1.2.0,<2;python_version<'3.10'",
+        "pandas>=1.3.0,<2;python_version>='3.10'",
+        'humanfriendly>=8.2,<11',
+        'pandas @ git+https://github.com/pandas-dev/pandas.git@master',
+    ]
+
+    # Run
+    minimum_versions_39 = _get_minimum_versions(dependencies, '3.9')
+    minimum_versions_310 = _get_minimum_versions(dependencies, '3.10')
+
+    # Assert
+    expected_versions_39 = {
+        'numpy': 'numpy==1.20.0',
+        'pandas': 'git+https://github.com/pandas-dev/pandas.git@master#egg=pandas',
+        'humanfriendly': 'humanfriendly==8.2',
+    }
+    expected_versions_310 = {
+        'numpy': 'numpy==1.23.3',
+        'pandas': 'git+https://github.com/pandas-dev/pandas.git@master#egg=pandas',
+        'humanfriendly': 'humanfriendly==8.2',
+    }
+
+    assert minimum_versions_39 == expected_versions_39
+    assert minimum_versions_310 == expected_versions_310
+
+
+def _get_example_pyproject_dict():
+    return {
+        'build-system': {
+            'build-backend': 'setuptools.build_meta',
+            'requires': ['setuptools', 'wheel'],
+        },
+        'project': {
+            'authors': [{'email': 'info@sdv.dev', 'name': 'DataCebo, Inc.'}],
+            'classifiers': [
+                'Intended Audience :: Developers',
+                'License :: Free for non-commercial use',
+                'Natural Language :: English',
+                'Programming Language :: Python :: 3.10',
+                'Programming Language :: Python :: 3.11',
+                'Programming Language :: Python :: 3.12',
+                'Topic :: Scientific/Engineering :: Artificial Intelligence',
+            ],
+            'dependencies': [
+                'appdirs>=1.3',
+                'boto3>=1.28,<2',
+                'botocore>=1.31,<2',
+                'cloudpickle>=2.1.0',
+                'compress-pickle>=1.2.0',
+                'humanfriendly>=8.2',
+                "numpy>=1.21.6;python_version<'3.10'",
+                "numpy>=1.23.3;python_version>='3.10' and python_version<'3.12'",
+                "numpy>=1.26.0;python_version>='3.12'",
+                "pandas>=1.4.0;python_version<'3.11'",
+                "pandas>=1.5.0;python_version>='3.11' and python_version<'3.12'",
+                "pandas>=2.1.1;python_version>='3.12'",
+                'psutil>=5.7',
+                "scikit-learn>=1.0.2;python_version<'3.10'",
+                "scikit-learn>=1.1.0;python_version>='3.10' and python_version<'3.11'",
+                "scikit-learn>=1.1.3;python_version>='3.11' and python_version<'3.12'",
+                "scikit-learn>=1.3.1;python_version>='3.12'",
+                "scipy>=1.7.3;python_version<'3.10'",
+                "scipy>=1.9.2;python_version>='3.10' and python_version<'3.12'",
+                "scipy>=1.12.0;python_version>='3.12'",
+                'tabulate>=0.8.3,<0.9',
+                "torch>=1.12.1;python_version<'3.10'",
+                "torch>=2.0.0;python_version>='3.10' and python_version<'3.12'",
+                "torch>=2.2.0;python_version>='3.12'",
+                'tqdm>=4.66.3',
+                'XlsxWriter>=1.2.8',
+                'rdt>=1.13.1',
+                'sdmetrics>=0.17.0',
+                'sdv>=1.17.2',
+            ],
+            'dynamic': ['version'],
+            'license': {'text': 'BSL-1.1'},
+            'name': 'sdgym',
+            'optional-dependencies': {
+                'all': ['sdgym[dask, test, dev]'],
+                'dask': ['dask', 'distributed'],
+                'dev': [
+                    'sdgym[dask, test]',
+                    'build>=1.0.0,<2',
+                    'bump-my-version>=0.18.3,<1',
+                    'pip>=9.0.1',
+                    'watchdog>=1.0.1,<5',
+                    'ruff>=0.4.5,<1',
+                    'twine>=1.10.0,<6',
+                    'wheel>=0.30.0',
+                    'coverage>=4.5.12,<8',
+                    'tox>=2.9.1,<5',
+                    'importlib-metadata>=3.6',
+                    'invoke',
+                ],
+                'realtabformer': ['realtabformer>=0.2.1'],
+                'test': [
+                    'sdgym[realtabformer]',
+                    'pytest>=6.2.5',
+                ],
+            },
+            'readme': 'README.md',
+            'requires-python': '>=3.8,<3.13',
+        },
+        'tool': {
+            'bumpversion': {
+                'allow_dirty': False,
+                'commit': True,
+                'commit_args': '',
+            },
+            'ruff': {
+                'exclude': [
+                    'docs',
+                    '.tox',
+                    '.git',
+                    '__pycache__',
+                    '.ipynb_checkpoints',
+                    'tasks.py',
+                ],
+                'indent-width': 4,
+            },
+        },
+    }
+
+
+def test__get_extra_dependencies():
+    """Test that the proper dependency strings are extracted from the pyproject dictionary."""
+    # Setup
+    pyproject_dict = _get_example_pyproject_dict()
+
+    # Run
+    extra_dependencies = _get_extra_dependencies(pyproject_dict)
+
+    # Assert
+    assert extra_dependencies == ['realtabformer>=0.2.1']
+
+
+def test__resolve_version_conflicts_conflicting_versions():
+    """Test that any conflicts for the same dependency are resolved to the higher version."""
+    # Setup
+    deps = {
+        'numpy': 'numpy==2.0.1',
+        'pandas': 'pandas==2.2.1',
+        'sdv': 'sdv==2.1.1',
+        'rdt': 'rdt==1.1.2',
+    }
+    extra_deps = {
+        'numpy': 'numpy==2.0.0',
+        'pandas': 'pandas==2.3.0',
+        'sdv': 'sdv==3.0.0',
+        'copulas': 'copulas==0.12.0',
+    }
+
+    # Run
+    versions = _resolve_version_conflicts(deps, extra_deps)
+
+    # Assert
+    assert sorted(versions) == sorted([
+        'numpy==2.0.1',
+        'pandas==2.3.0',
+        'sdv==3.0.0',
+        'rdt==1.1.2',
+        'copulas==0.12.0',
+    ])
+
+
+def test__resolve_version_conflicts_pointing_to_branch():
+    """Test specific branches are always selected over normal version numbers."""
+    # Setup
+    deps = {
+        'numpy': 'git+https://github.com/numpy-dev/numpy.git@master#egg=numpy',
+        'pandas': 'pandas==2.2.1',
+        'sdv': 'sdv==2.1.1',
+        'rdt': 'rdt==1.1.2',
+    }
+    extra_deps = {
+        'numpy': 'numpy==2.0.0',
+        'pandas': 'git+https://github.com/pandas-dev/pandas.git@master#egg=pandas',
+        'sdv': 'sdv==3.0.0',
+        'copulas': 'copulas==0.12.0',
+    }
+
+    # Run
+    versions = _resolve_version_conflicts(deps, extra_deps)
+
+    # Assert
+    assert sorted(versions) == sorted([
+        'git+https://github.com/numpy-dev/numpy.git@master#egg=numpy',
+        'git+https://github.com/pandas-dev/pandas.git@master#egg=pandas',
+        'sdv==3.0.0',
+        'rdt==1.1.2',
+        'copulas==0.12.0',
+    ])

sdgym-0.9.1.dev0/tests/test_tasks.py

@@ -1,39 +0,0 @@
-"""Tests for the ``tasks.py`` file."""
-
-from tasks import _get_minimum_versions
-
-
-def test_get_minimum_versions():
-    """Test the ``_get_minimum_versions`` method.
-
-    The method should return the minimum versions of the dependencies for the given python version.
-    If a library is linked to an URL, the minimum version should be the URL.
-    """
-    # Setup
-    dependencies = [
-        "numpy>=1.20.0,<2;python_version<'3.10'",
-        "numpy>=1.23.3,<2;python_version>='3.10'",
-        "pandas>=1.2.0,<2;python_version<'3.10'",
-        "pandas>=1.3.0,<2;python_version>='3.10'",
-        'humanfriendly>=8.2,<11',
-        'pandas @ git+https://github.com/pandas-dev/pandas.git@master',
-    ]
-
-    # Run
-    minimum_versions_39 = _get_minimum_versions(dependencies, '3.9')
-    minimum_versions_310 = _get_minimum_versions(dependencies, '3.10')
-
-    # Assert
-    expected_versions_39 = [
-        'numpy==1.20.0',
-        'git+https://github.com/pandas-dev/pandas.git@master#egg=pandas',
-        'humanfriendly==8.2',
-    ]
-    expected_versions_310 = [
-        'numpy==1.23.3',
-        'git+https://github.com/pandas-dev/pandas.git@master#egg=pandas',
-        'humanfriendly==8.2',
-    ]
-
-    assert minimum_versions_39 == expected_versions_39
-    assert minimum_versions_310 == expected_versions_310