Replace MultiTableMetadata with Metadata

sdv/io/local/local.py

@@ -7,7 +7,7 @@
 
 import pandas as pd
 
-from sdv.metadata import MultiTableMetadata
+from sdv.metadata import Metadata
 
 
 class BaseLocalHandler:
@@ -25,11 +25,11 @@ def create_metadata(self, data):
                 Dictionary of table names to dataframes.
 
         Returns:
-            MultiTableMetadata:
-                An ``sdv.metadata.MultiTableMetadata`` object with the detected metadata
+            Metadata:
+                An ``sdv.metadata.Metadata`` object with the detected metadata
                 properties from the data.
         """
-        metadata = MultiTableMetadata()
+        metadata = Metadata()
         metadata.detect_from_dataframes(data)
         return metadata
 

sdv/metadata/multi_table.py

@@ -26,6 +26,10 @@
 LOGGER = logging.getLogger(__name__)
 MULTITABLEMETADATA_LOGGER = get_sdv_logger('MultiTableMetadata')
 WARNINGS_COLUMN_ORDER = ['Table Name', 'Column Name', 'sdtype', 'datetime_format']
+DEPRECATION_MSG = (
+    "The 'MultiTableMetadata' is deprecated. Please use the new "
+    "'Metadata' class for synthesizers."
+)
 
 
 class MultiTableMetadata:

sdv/metadata/single_table.py

@@ -1371,7 +1371,7 @@ def upgrade_metadata(cls, filepath):
             if len(tables) > 1:
                 raise InvalidMetadataError(
                     'There are multiple tables specified in the JSON. '
-                    'Try using the MultiTableMetadata class to upgrade this file.'
+                    'Try using the Metadata class to upgrade this file.'
                 )
 
             else:

sdv/multi_table/base.py

@@ -26,6 +26,8 @@
     SynthesizerInputError,
 )
 from sdv.logging import disable_single_table_logger, get_sdv_logger
+from sdv.metadata.metadata import Metadata
+from sdv.metadata.multi_table import DEPRECATION_MSG, MultiTableMetadata
 from sdv.single_table.copulas import GaussianCopulaSynthesizer
 
 SYNTHESIZER_LOGGER = get_sdv_logger('MultiTableSynthesizer')
@@ -38,8 +40,8 @@ class BaseMultiTableSynthesizer:
     multi table synthesizers need to implement, as well as common functionality.
 
     Args:
-        metadata (sdv.metadata.multi_table.MultiTableMetadata):
-            Multi table metadata representing the data tables that this synthesizer will be used
+        metadata (sdv.metadata.Metadata):
+            Table 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.
@@ -71,8 +73,9 @@ def _initialize_models(self):
         with disable_single_table_logger():
             for table_name, table_metadata in self.metadata.tables.items():
                 synthesizer_parameters = self._table_parameters.get(table_name, {})
+                metadata = Metadata.load_from_dict(table_metadata.to_dict())
                 self._table_synthesizers[table_name] = self._synthesizer(
-                    metadata=table_metadata, locales=self.locales, **synthesizer_parameters
+                    metadata=metadata, locales=self.locales, **synthesizer_parameters
                 )
                 self._table_synthesizers[table_name]._data_processor.table_name = table_name
 
@@ -97,6 +100,9 @@ def _check_metadata_updated(self):
 
     def __init__(self, metadata, locales=['en_US'], synthesizer_kwargs=None):
         self.metadata = metadata
+        if type(metadata) is MultiTableMetadata:
+            self.metadata = Metadata().load_from_dict(metadata.to_dict())
+            warnings.warn(DEPRECATION_MSG, FutureWarning)
         with warnings.catch_warnings():
             warnings.filterwarnings('ignore', message=r'.*column relationship.*')
             self.metadata.validate()

sdv/multi_table/hma.py

@@ -23,7 +23,7 @@ class HMASynthesizer(BaseHierarchicalSampler, BaseMultiTableSynthesizer):
     """Hierarchical Modeling Algorithm One.
 
     Args:
-        metadata (sdv.metadata.multi_table.MultiTableMetadata):
+        metadata (sdv.metadata.Metadata):
             Multi table metadata representing the data tables that this synthesizer will be used
             for.
         locales (list or str):
@@ -47,7 +47,7 @@ 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 (Metadata):
                 Metadata of the datasets.
         """
         columns_per_table = {}

sdv/multi_table/utils.py

@@ -123,7 +123,7 @@ def _simplify_relationships_and_tables(metadata, tables_to_drop):
     Removes the tables that are not direct child or grandchild of the root table.
 
     Args:
-        metadata (MultiTableMetadata):
+        metadata (Metadata):
             Metadata of the datasets.
         tables_to_drop (set):
             Set of the tables that relationships will be removed.
@@ -149,7 +149,7 @@ def _simplify_grandchildren(metadata, grandchildren):
      - Drop all modelables columns.
 
     Args:
-        metadata (MultiTableMetadata):
+        metadata (Metadata):
             Metadata of the datasets.
         grandchildren (set):
             Set of the grandchildren of the root table.
@@ -174,7 +174,7 @@ def _get_num_column_to_drop(metadata, child_table, max_col_per_relationships):
         - minimum number of column to drop = n + k - sqrt(k^2 + 1 + 2m)
 
     Args:
-        metadata (MultiTableMetadata):
+        metadata (Metadata):
             Metadata of the datasets.
         child_table (str):
             Name of the child table.
@@ -232,7 +232,7 @@ def _simplify_child(metadata, child_table, max_col_per_relationships):
     """Simplify the child table.
 
     Args:
-        metadata (MultiTableMetadata):
+        metadata (Metadata):
             Metadata of the datasets.
         child_table (str):
             Name of the child table.
@@ -252,7 +252,7 @@ def _simplify_children(metadata, children, root_table, num_data_column):
      - Drop some modelable columns to have at most MAX_NUMBER_OF_COLUMNS columns to model.
 
     Args:
-        metadata (MultiTableMetadata):
+        metadata (Metadata):
             Metadata of the datasets.
         children (set):
             Set of the children of the root table.
@@ -288,11 +288,11 @@ def _simplify_metadata(metadata):
      - Drop some modelable columns in the children to have at most 1000 columns to model.
 
     Args:
-        metadata (MultiTableMetadata):
+        metadata (Metadata):
             Metadata of the datasets.
 
     Returns:
-        MultiTableMetadata:
+        Metadata:
             Simplified metadata.
     """
     simplified_metadata = deepcopy(metadata)
@@ -330,7 +330,7 @@ def _simplify_data(data, metadata):
         data (dict):
             Dictionary that maps each table name (string) to the data for that
             table (pandas.DataFrame).
-        metadata (MultiTableMetadata):
+        metadata (Metadata):
             Metadata of the datasets.
 
     Returns:
@@ -375,7 +375,7 @@ def _get_rows_to_drop(data, metadata):
     This ensures that we preserve the referential integrity between all the relationships.
 
     Args:
-        metadata (MultiTableMetadata):
+        metadata (Metadata):
             Metadata of the datasets.
         data (dict):
             Dictionary that maps each table name (string) to the data for that
@@ -470,7 +470,7 @@ def _subsample_table_and_descendants(data, metadata, table, num_rows):
         data (dict):
             Dictionary that maps each table name (string) to the data for that
             table (pandas.DataFrame).
-        metadata (MultiTableMetadata):
+        metadata (Metadata):
             Metadata of the datasets.
         table (str):
             Name of the table.
@@ -496,7 +496,7 @@ def _get_primary_keys_referenced(data, metadata):
         data (dict):
             Dictionary that maps each table name (string) to the data for that
             table (pandas.DataFrame).
-        metadata (MultiTableMetadata):
+        metadata (Metadata):
             Metadata of the datasets.
 
     Returns:
@@ -568,7 +568,7 @@ def _subsample_ancestors(data, metadata, table, primary_keys_referenced):
         data (dict):
             Dictionary that maps each table name (string) to the data for that
             table (pandas.DataFrame).
-        metadata (MultiTableMetadata):
+        metadata (Metadata):
             Metadata of the datasets.
         table (str):
             Name of the table.
@@ -604,7 +604,7 @@ def _subsample_data(data, metadata, main_table_name, num_rows):
       referenced by the descendants and some unreferenced rows.
 
     Args:
-        metadata (MultiTableMetadata):
+        metadata (Metadata):
             Metadata of the datasets.
         data (dict):
             Dictionary that maps each table name (string) to the data for that

sdv/sampling/independent_sampler.py

@@ -10,7 +10,7 @@ class BaseIndependentSampler:
     """Independent sampler mixin.
 
     Args:
-        metadata (sdv.metadata.multi_table.MultiTableMetadata):
+        metadata (sdv.metadata.Metadata):
             Multi-table metadata representing the data tables that this sampler will be used for.
         table_synthesizers (dict):
             Dictionary mapping each table to a synthesizer. Should be instantiated and passed to

sdv/utils/poc.py

@@ -40,7 +40,7 @@ def simplify_schema(data, metadata, verbose=True):
         data (dict):
             Dictionary that maps each table name (string) to the data for that
             table (pandas.DataFrame).
-        metadata (MultiTableMetadata):
+        metadata (Metadata):
             Metadata of the datasets.
         verbose (bool):
             If True, print information about the simplification process.
@@ -50,7 +50,7 @@ def simplify_schema(data, metadata, verbose=True):
         tuple:
             dict:
                 Dictionary with the simplified dataframes.
-            MultiTableMetadata:
+            Metadata:
                 Simplified metadata.
     """
     try:
@@ -93,7 +93,7 @@ def get_random_subset(data, metadata, main_table_name, num_rows, verbose=True):
         data (dict):
             Dictionary that maps each table name (string) to the data for that
             table (pandas.DataFrame).
-        metadata (MultiTableMetadata):
+        metadata (Metadata):
             Metadata of the datasets.
         main_table_name (str):
             Name of the main table.

sdv/utils/utils.py

@@ -18,7 +18,7 @@ def drop_unknown_references(data, metadata, drop_missing_values=True, verbose=Tr
         data (dict):
             Dictionary that maps each table name (string) to the data for that
             table (pandas.DataFrame).
-        metadata (MultiTableMetadata):
+        metadata (Metadata):
             Metadata of the datasets.
         drop_missing_values (bool):
             Boolean describing whether or not to also drop foreign keys with missing values

tests/integration/metadata/test_metadata.py

@@ -1,8 +1,11 @@
+import re
+
 import pytest
 
 from sdv.datasets.demo import download_demo
 from sdv.metadata.metadata import DEFAULT_TABLE_NAME, Metadata
 from sdv.metadata.multi_table import MultiTableMetadata
+from sdv.multi_table.hma import HMASynthesizer
 from sdv.single_table.copulas import GaussianCopulaSynthesizer
 
 
@@ -255,3 +258,41 @@ def test_single_table_compatibility(tmp_path):
     metadata_sample = synthesizer.sample(10)
     assert loaded_synthesizer.metadata.to_dict() == synthesizer_2.metadata.to_dict()
     assert metadata_sample.columns.to_list() == loaded_sample.columns.to_list()
+
+
+def test_multi_table_compatibility(tmp_path):
+    """Test if SingleMetadataTable still has compatibility with single table synthesizers."""
+    # Setup
+    data, metadata = download_demo('multi_table', 'fake_hotels')
+    warn_msg = re.escape(
+        "The 'MultiTableMetadata' is deprecated. Please use the new "
+        "'Metadata' class for synthesizers."
+    )
+
+    # Run
+    with pytest.warns(FutureWarning, match=warn_msg):
+        synthesizer = HMASynthesizer(metadata)
+    synthesizer.fit(data)
+    model_path = tmp_path / 'synthesizer.pkl'
+    synthesizer.save(model_path)
+
+    # Assert
+    assert model_path.exists()
+    assert model_path.is_file()
+    loaded_synthesizer = HMASynthesizer.load(model_path)
+    assert isinstance(synthesizer, HMASynthesizer)
+    assert loaded_synthesizer.get_info() == synthesizer.get_info()
+    assert isinstance(loaded_synthesizer.metadata, Metadata)
+    expected_metadata = metadata.to_dict()
+    expected_metadata['METADATA_SPEC_VERSION'] = 'V1'
+    assert loaded_synthesizer.metadata.to_dict() == expected_metadata
+    loaded_sample = loaded_synthesizer.sample(10)
+    synthesizer.validate(loaded_sample)
+
+    # Run against Metadata
+    synthesizer_2 = HMASynthesizer(Metadata._convert_to_unified_metadata(metadata))
+    synthesizer_2.fit(data)
+    metadata_sample = synthesizer.sample(10)
+    assert loaded_synthesizer.metadata.to_dict() == synthesizer_2.metadata.to_dict()
+    for table in metadata_sample:
+        assert metadata_sample[table].columns.to_list() == loaded_sample[table].columns.to_list()

tests/integration/multi_table/test_hma.py

@@ -18,6 +18,7 @@
 from sdv.datasets.local import load_csvs
 from sdv.errors import SamplingError, SynthesizerInputError, VersionError
 from sdv.evaluation.multi_table import evaluate_quality, get_column_pair_plot, get_column_plot
+from sdv.metadata.metadata import Metadata
 from sdv.metadata.multi_table import MultiTableMetadata
 from sdv.multi_table import HMASynthesizer
 from tests.integration.single_table.custom_constraints import MyConstraint
@@ -393,6 +394,7 @@ def test_save_and_load(self, tmp_path):
         """Test saving and loading a multi-table synthesizer."""
         # Setup
         _, _, metadata = self.get_custom_constraint_data_and_metadata()
+        metadata = Metadata.load_from_dict(metadata.to_dict())
         synthesizer = HMASynthesizer(metadata)
         model_path = tmp_path / 'synthesizer.pkl'
 
@@ -460,6 +462,7 @@ def test_synthesize_multiple_tables_using_hma(self, tmp_path):
         """
         # Loading the demo data
         real_data, metadata = download_demo(modality='multi_table', dataset_name='fake_hotels')
+        metadata = Metadata.load_from_dict(metadata.to_dict())
 
         # Creating a Synthesizer
         synthesizer = HMASynthesizer(metadata)
@@ -1250,9 +1253,8 @@ def test_metadata_updated_no_warning(self, tmp_path):
             instance.fit(data)
 
         # Assert
-        for warning in captured_warnings:
-            assert warning.category is FutureWarning
-        assert len(captured_warnings) == 3
+        assert len(captured_warnings) == 1
+        assert captured_warnings[0].category is FutureWarning
 
         # Run 2
         metadata_detect = MultiTableMetadata()
@@ -1271,9 +1273,8 @@ def test_metadata_updated_no_warning(self, tmp_path):
             instance.fit(data)
 
         # Assert
-        for warning in captured_warnings:
-            assert warning.category is FutureWarning
-        assert len(captured_warnings) == 3
+        assert len(captured_warnings) == 1
+        assert captured_warnings[0].category is FutureWarning
 
         # Run 3
         instance = HMASynthesizer(metadata_detect)
@@ -1297,7 +1298,7 @@ def test_metadata_updated_warning_detect(self):
         """
         # Setup
         data, metadata = download_demo('multi_table', 'got_families')
-        metadata_detect = MultiTableMetadata()
+        metadata_detect = Metadata()
         metadata_detect.detect_from_dataframes(data)
 
         metadata_detect.relationships = metadata.relationships
@@ -1316,16 +1317,7 @@ def test_metadata_updated_warning_detect(self):
             instance.fit(data)
 
         # Assert
-        future_warnings = 0
-        user_warnings = 0
-        for warning in record:
-            if warning.category is FutureWarning:
-                future_warnings += 1
-            if warning.category is UserWarning:
-                user_warnings += 1
-        assert future_warnings == 3
-        assert user_warnings == 1
-        assert len(record) == 4
+        assert len(record) == 1
 
     def test_null_foreign_keys(self):
         """Test that the synthesizer crashes when there are null foreign keys."""
@@ -1595,7 +1587,7 @@ def test_metadata_updated_warning(method, kwargs):
 
     The warning should be raised during synthesizer initialization.
     """
-    metadata = MultiTableMetadata().load_from_dict({
+    metadata = Metadata().load_from_dict({
         'tables': {
             'departure': {
                 'primary_key': 'id',