make release-tag: Merge branch 'main' into stable

.github/workflows/dtypes_benchmark.yml

@@ -8,6 +8,7 @@ on:
 jobs:
   run_dtypes_benchmark:
     runs-on: ubuntu-latest
+    timeout-minutes: 20
 
     strategy:
       matrix:
@@ -39,6 +40,8 @@ jobs:
         run: |
           invoke benchmark-dtypes
 
+        continue-on-error: true
+
       # Upload the json files as artifacts
       - name: Upload artifacts
         uses: actions/upload-artifact@v3

.github/workflows/static_code_analysis.yml

@@ -24,7 +24,7 @@ jobs:
       id: cpr
       uses: peter-evans/create-pull-request@v4
       with:
-        token: ${{ secrets.GITHUB_TOKEN }}
+        token: ${{ secrets.GITHUB_ACCESS_TOKEN }}
         commit-message: Update static code analysis
         author: "github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>"
         committer: "github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>"

HISTORY.md

@@ -1,6 +1,21 @@
 # Release Notes
 
-### v1.17.0 - 2024-10-02
+### v1.17.1 - 2024-10-10
+
+### Bugs Fixed
+
+* Extraneous metadata warning is printed when customizing HMA Synthesizer - Issue [#2250](https://github.com/sdv-dev/SDV/issues/2250) by @pvk-developer
+
+### Internal
+
+* `ValueError` raised when adding columns to a new `Metadata` instance - Issue [#2252](https://github.com/sdv-dev/SDV/issues/2252) by @pvk-developer
+* Enhance Benchmark Report Generation - Issue [#2235](https://github.com/sdv-dev/SDV/issues/2235) by @pvk-developer
+
+### Maintenance
+
+* Use PyDrive2 instead of PyDrive - Issue [#2238](https://github.com/sdv-dev/SDV/issues/2238) by @pvk-developer
+
+## v1.17.0 - 2024-10-02
 
 This release consolidates the `SingleTableMetadata` and `MultiTableMetadata` classes into one new class simply called `Metadata`. The old classes are now deprecated. The new class has the same structure as the `MultiTableMetadata` class, except it will work with single table synthesizers.
 

pyproject.toml

@@ -62,7 +62,10 @@ test = [
     'rundoc>=0.4.3,<0.5',
     'pytest-runner >= 2.11.1',
     'tomli>=2.0.0,<3',
-    'pydrive',
+    'google-api-python-client',
+    'google-auth',
+    'google-auth-oauthlib',
+    'requests',
     'pyarrow',
     'gitpython',
     'slack-sdk>=3.23,<4.0',
@@ -136,7 +139,7 @@ namespaces = false
 version = {attr = 'sdv.__version__'}
 
 [tool.bumpversion]
-current_version = "1.17.0"
+current_version = "1.17.1.dev1"
 parse = '(?P<major>\d+)\.(?P<minor>\d+)\.(?P<patch>\d+)(\.(?P<release>[a-z]+)(?P<candidate>\d+))?'
 serialize = [
     '{major}.{minor}.{patch}.{release}{candidate}',

sdv/__init__.py

@@ -6,7 +6,7 @@
 
 __author__ = 'DataCebo, Inc.'
 __email__ = 'info@sdv.dev'
-__version__ = '1.17.0'
+__version__ = '1.17.1.dev1'
 
 
 import sys

sdv/metadata/metadata.py

@@ -131,6 +131,7 @@ def _set_metadata_dict(self, metadata, single_table_name=None):
                     'No table name was provided to metadata containing only one table. '
                     f'Assigning name: {single_table_name}'
                 )
+
             self.tables[single_table_name] = SingleTableMetadata.load_from_dict(metadata)
 
     def _get_single_table_name(self):
@@ -162,6 +163,8 @@ def _convert_to_single_table(self):
         return next(iter(self.tables.values()), SingleTableMetadata())
 
     def _handle_table_name(self, table_name):
+        if len(self.tables) == 0:
+            raise ValueError('Metadata does not contain any tables. No columns can be added.')
         if table_name is None:
             if len(self.tables) == 1:
                 table_name = next(iter(self.tables))
@@ -270,3 +273,18 @@ def add_alternate_keys(self, column_names, table_name=None):
         """Add alternate keys to a table."""
         table_name = self._handle_table_name(table_name)
         super().add_alternate_keys(table_name, column_names)
+
+    def get_table_metadata(self, table_name=None):
+        """Return the metadata for a table.
+
+        Args:
+            table_name (str):
+                The name of the table to get the metadata for.
+
+        Returns:
+            Metadata:
+                The metadata for the given table.
+        """
+        table_name = self._handle_table_name(table_name)
+        table_metadata = super().get_table_metadata(table_name)
+        return Metadata.load_from_dict(table_metadata.to_dict(), single_table_name=table_name)

sdv/multi_table/base.py

@@ -209,8 +209,9 @@ def set_table_parameters(self, table_name, table_parameters):
                 A dictionary with the parameters as keys and the values to be used to instantiate
                 the table's synthesizer.
         """
+        table_metadata = self.metadata.get_table_metadata(table_name)
         self._table_synthesizers[table_name] = self._synthesizer(
-            metadata=self.metadata.tables[table_name], **table_parameters
+            metadata=table_metadata, **table_parameters
         )
         self._table_parameters[table_name].update(deepcopy(table_parameters))
 

sdv/multi_table/hma.py

@@ -23,9 +23,8 @@ class HMASynthesizer(BaseHierarchicalSampler, BaseMultiTableSynthesizer):
     """Hierarchical Modeling Algorithm One.
 
     Args:
-        metadata (sdv.metadata.multi_table.MultiTableMetadata):
-            Multi table metadata representing the data tables that this synthesizer will be used
-            for.
+        metadata (sdv.metadata.Metadata):
+            Metadata representing the data tables that this synthesizer will be used for.
         locales (list or str):
             The default locale(s) to use for AnonymizedFaker transformers.
             Defaults to ``['en_US']``.
@@ -47,8 +46,8 @@ def _get_num_data_columns(metadata):
         """Get the number of data columns, ie colums that are not id, for each table.
 
         Args:
-            metadata (MultiTableMetadata):
-                Metadata of the datasets.
+            metadata (sdv.metadata.Metadata):
+                Metadata representing the data tables that this synthesizer will be used for.
         """
         columns_per_table = {}
         for table_name, table in metadata.tables.items():
@@ -552,8 +551,7 @@ def _recreate_child_synthesizer(self, child_name, parent_name, parent_row):
         if parent_row is not None:
             parameters = self._extract_parameters(parent_row, child_name, foreign_key)
             default_parameters = getattr(self, '_default_parameters', {}).get(child_name, {})
-
-            table_meta = self.metadata.tables[child_name]
+            table_meta = self.metadata.get_table_metadata(child_name)
             synthesizer = self._synthesizer(table_meta, **self._table_parameters[child_name])
             synthesizer._set_parameters(parameters, default_parameters)
         else:

sdv/sampling/hierarchical_sampler.py

@@ -95,13 +95,10 @@ def _add_child_rows(self, child_name, parent_name, parent_row, sampled_data, num
         foreign_key = self.metadata._get_foreign_keys(parent_name, child_name)[0]
         if num_rows is None:
             num_rows = parent_row[f'__{child_name}__{foreign_key}__num_rows']
-        with warnings.catch_warnings():
-            warnings.filterwarnings('ignore', message=".*The 'SingleTableMetadata' is deprecated.*")
-            child_synthesizer = self._recreate_child_synthesizer(
-                child_name, parent_name, parent_row
-            )
-        sampled_rows = self._sample_rows(child_synthesizer, num_rows)
 
+        child_synthesizer = self._recreate_child_synthesizer(child_name, parent_name, parent_row)
+
+        sampled_rows = self._sample_rows(child_synthesizer, num_rows)
         if len(sampled_rows):
             parent_key = self.metadata.tables[parent_name].primary_key
             if foreign_key in sampled_rows:

static_code_analysis.txt

@@ -1,4 +1,4 @@
-Run started:2024-09-26 20:22:23.878117
+Run started:2024-10-02 16:26:36.866545
 
 Test results:
 >> Issue: [B105:hardcoded_password_string] Possible hardcoded password: '# Release Notes
@@ -69,7 +69,7 @@ Test results:
 --------------------------------------------------
 
 Code scanned:
-	Total lines of code: 12468
+	Total lines of code: 12759
 	Total lines skipped (#nosec): 0
 	Total potential issues skipped due to specifically being disabled (e.g., #nosec BXXX): 0