Create PARSynthesizer (#1068)

* initial commit

* fixing skeleton for parsynthesizer

* adding unit tests

* pr comments

sdv/sequential/__init__.py

@@ -0,0 +1 @@
+"""Synthesizers for sequential data."""

sdv/sequential/par.py

@@ -0,0 +1,93 @@
+"""PAR Synthesizer class."""
+
+import inspect
+
+from sdv.data_processing import DataProcessor
+from sdv.metadata.single_table import SingleTableMetadata
+from sdv.single_table import GaussianCopulaSynthesizer
+
+
+class PARSynthesizer:
+    """Synthesizer for sequential data.
+
+    This synthesizer uses the ``deepecho.models.par.PARModel`` class as the core model.
+    Additionally, it uses a separate synthesizer to model and sample the context columns
+    to be passed into PAR.
+
+    Args:
+        metadata (sdv.metadata.SingleTableMetadata):
+            Single table metadata representing the data that this synthesizer will be used for.
+        enforce_min_max_values (bool):
+            Specify whether or not to clip the data returned by ``reverse_transform`` of
+            the numerical transformer, ``FloatFormatter``, to the min and max values seen
+            during ``fit``. Defaults to ``True``.
+        enforce_rounding (bool):
+            Define rounding scheme for ``numerical`` columns. If ``True``, the data returned
+            by ``reverse_transform`` will be rounded as in the original data. Defaults to ``True``.
+        context_columns (list[str]):
+            A list of strings, representing the columns that do not vary in a sequence.
+        segment_size (int):
+            If specified, cut each training sequence in several segments of
+            the indicated size. The size can be passed as an integer
+            value, which will interpreted as the number of data points to
+            put on each segment.
+        epochs (int):
+            The number of epochs to train for. Defaults to 128.
+        sample_size (int):
+            The number of times to sample (before choosing and
+            returning the sample which maximizes the likelihood).
+            Defaults to 1.
+        cuda (bool):
+            Whether to attempt to use cuda for GPU computation.
+            If this is False or CUDA is not available, CPU will be used.
+            Defaults to ``True``.
+        verbose (bool):
+            Whether to print progress to console or not.
+    """
+
+    def _get_context_metadata(self):
+        context_columns_dict = {}
+        context_columns = self.context_columns or []
+        for column in context_columns:
+            context_columns_dict[column] = self.metadata._columns[column]
+
+        context_metadata_dict = {'columns': context_columns_dict}
+        return SingleTableMetadata._load_from_dict(context_metadata_dict)
+
+    def __init__(self, metadata, enforce_min_max_values, enforce_rounding, context_columns=None,
+                 segment_size=None, epochs=128, sample_size=1, cuda=True, verbose=False):
+        self.metadata = metadata
+        self.enforce_min_max_values = enforce_min_max_values
+        self.enforce_rounding = enforce_rounding
+        self._data_processor = DataProcessor(metadata)
+        self.context_columns = context_columns
+        self.segment_size = segment_size
+        self._model_kwargs = {
+            'epochs': epochs,
+            'sample_size': sample_size,
+            'cuda': cuda,
+            'verbose': verbose,
+        }
+        context_metadata = self._get_context_metadata()
+        self._context_synthesizer = GaussianCopulaSynthesizer(
+            metadata=context_metadata,
+            enforce_min_max_values=enforce_min_max_values,
+            enforce_rounding=enforce_rounding
+        )
+
+    def get_parameters(self):
+        """Return the parameters used to instantiate the synthesizer."""
+        parameters = inspect.signature(self.__init__).parameters
+        instantiated_parameters = {}
+        for parameter_name in parameters:
+            if parameter_name != 'metadata':
+                instantiated_parameters[parameter_name] = self.__dict__.get(parameter_name)
+
+        for parameter_name, value in self._model_kwargs.items():
+            instantiated_parameters[parameter_name] = value
+
+        return instantiated_parameters
+
+    def get_metadata(self):
+        """Return the ``SingleTableMetadata`` for this synthesizer."""
+        return self.metadata

sdv/single_table/__init__.py

@@ -1,4 +1,4 @@
-"""Models for Single Table data."""
+"""Synthesizers for Single Table data."""
 
 from sdv.single_table.copulagan import CopulaGANSynthesizer
 from sdv.single_table.copulas import GaussianCopulaSynthesizer

tests/unit/sequential/__init__.py

@@ -0,0 +1 @@
+"""Tests for synthesizers for sequential data."""

tests/unit/sequential/test_par.py

@@ -0,0 +1,106 @@
+from sdv.data_processing.data_processor import DataProcessor
+from sdv.metadata.single_table import SingleTableMetadata
+from sdv.sequential.par import PARSynthesizer
+from sdv.single_table.copulas import GaussianCopulaSynthesizer
+
+
+class TestPARSynthesizer:
+
+    def test___init__(self):
+        """Test that the parameters are set correctly.
+
+        The parameters passed in the ``__init__`` should be set on the instance. Additionally,
+        a context synthesizer should be created with the correct metadata and parameters.
+        """
+        # Setup
+        metadata = SingleTableMetadata()
+        metadata.add_column('time', sdtype='datetime')
+        metadata.add_column('gender', sdtype='categorical')
+        metadata.add_column('name', sdtype='text')
+        metadata.add_column('measurement', sdtype='numerical')
+
+        # Run
+        synthesizer = PARSynthesizer(
+            metadata=metadata,
+            enforce_min_max_values=True,
+            enforce_rounding=True,
+            context_columns=['gender', 'name'],
+            segment_size=10,
+            epochs=10,
+            sample_size=5,
+            cuda=False,
+            verbose=False
+        )
+
+        # Assert
+        assert synthesizer.context_columns == ['gender', 'name']
+        assert synthesizer.enforce_min_max_values is True
+        assert synthesizer.enforce_rounding is True
+        assert synthesizer.segment_size == 10
+        assert synthesizer._model_kwargs == {
+            'epochs': 10,
+            'sample_size': 5,
+            'cuda': False,
+            'verbose': False
+        }
+        assert isinstance(synthesizer._data_processor, DataProcessor)
+        assert synthesizer._data_processor.metadata == metadata
+        assert isinstance(synthesizer._context_synthesizer, GaussianCopulaSynthesizer)
+        assert synthesizer._context_synthesizer.metadata._columns == {
+            'gender': {'sdtype': 'categorical'},
+            'name': {'sdtype': 'text'}
+        }
+
+    def test_get_parameters(self):
+        """Test that it returns every ``init`` parameter without the ``metadata``."""
+        # Setup
+        metadata = SingleTableMetadata()
+        instance = PARSynthesizer(
+            metadata=metadata,
+            enforce_min_max_values=True,
+            enforce_rounding=True,
+            context_columns=None,
+            segment_size=10,
+            epochs=10,
+            sample_size=5,
+            cuda=False,
+            verbose=False
+        )
+
+        # Run
+        parameters = instance.get_parameters()
+
+        # Assert
+        assert 'metadata' not in parameters
+        assert parameters == {
+            'enforce_min_max_values': True,
+            'enforce_rounding': True,
+            'context_columns': None,
+            'segment_size': 10,
+            'epochs': 10,
+            'sample_size': 5,
+            'cuda': False,
+            'verbose': False
+        }
+
+    def test_get_metadata(self):
+        """Test that it returns the ``metadata`` object."""
+        # Setup
+        metadata = SingleTableMetadata()
+        instance = PARSynthesizer(
+            metadata=metadata,
+            enforce_min_max_values=True,
+            enforce_rounding=True,
+            context_columns=None,
+            segment_size=10,
+            epochs=10,
+            sample_size=5,
+            cuda=False,
+            verbose=False
+        )
+
+        # Run
+        result = instance.get_metadata()
+
+        # Assert
+        assert result == metadata