Add plot column pair utility method

sdmetrics/reports/__init__.py

@@ -1,6 +1,7 @@
 """Reports for sdmetrics."""
-from sdmetrics.reports.utils import get_column_plot
+from sdmetrics.reports.utils import get_column_pair_plot, get_column_plot
 
 __all__ = [
+    'get_column_pair_plot',
     'get_column_plot',
 ]

sdmetrics/reports/utils.py

@@ -7,9 +7,15 @@
 import pandas as pd
 import plotly.express as px
 import plotly.figure_factory as ff
+from pandas.core.tools.datetimes import _guess_datetime_format_for_array
+
+from sdmetrics.utils import is_datetime
 
 DATACEBO_DARK = '#000036'
 DATACEBO_LIGHT = '#01E0C9'
+BACKGROUND_COLOR = '#F5F5F8'
+CONTINUOUS_SDTYPES = ['numerical', 'datetime']
+DISCRETE_SDTYPES = ['categorical', 'boolean']
 
 
 def make_discrete_column_plot(real_column, synthetic_column, sdtype):
@@ -79,7 +85,7 @@ def make_discrete_column_plot(real_column, synthetic_column, sdtype):
         title=f"Real vs. Synthetic Data for column '{column_name}'",
         xaxis_title='Category*' if show_missing_values else 'Category',
         yaxis_title='Frequency',
-        plot_bgcolor='#F5F5F8',
+        plot_bgcolor=BACKGROUND_COLOR,
         annotations=annotations,
     )
 
@@ -153,7 +159,7 @@ def make_continuous_column_plot(real_column, synthetic_column, sdtype):
         title=f'Real vs. Synthetic Data for column {column_name}',
         xaxis_title='Value*' if show_missing_values else 'Value',
         yaxis_title='Frequency',
-        plot_bgcolor='#F5F5F8',
+        plot_bgcolor=BACKGROUND_COLOR,
         annotations=annotations,
     )
 
@@ -169,19 +175,183 @@ def get_column_plot(real_column, synthetic_column, sdtype):
         synthetic_column (pandas.Series):
             The synthetic data for the desired column.
         sdtype (str):
-            The data type of the column.
+            The data type of the column. Must be one of
+            ('numerical', 'datetime', 'categorical', or 'boolean').
 
     Returns:
         plotly.graph_objects._figure.Figure
     """
-    if sdtype == 'numerical' or sdtype == 'datetime':
+    if sdtype in CONTINUOUS_SDTYPES:
         fig = make_continuous_column_plot(real_column, synthetic_column, sdtype)
-    elif sdtype == 'categorical' or sdtype == 'boolean':
+    elif sdtype in DISCRETE_SDTYPES:
         fig = make_discrete_column_plot(real_column, synthetic_column, sdtype)
+    else:
+        raise ValueError(f'sdtype of {sdtype} not recognized.')
+
+    return fig
+
+
+def convert_to_datetime(column_data):
+    """Convert a column data to pandas datetime.
+
+    Args:
+        column_data (pandas.Series):
+            The column data
+
+    Returns:
+        pandas.Series:
+            The converted column data.
+    """
+    if is_datetime(column_data):
+        return column_data
+
+    dt_format = _guess_datetime_format_for_array(column_data.astype(str).to_numpy())
+    return pd.to_datetime(column_data, format=dt_format)
+
+
+def make_continuous_column_pair_plot(real_data, synthetic_data):
+    """Make a column pair plot for continuous data.
+
+    Args:
+        real_data (pandas.DataFrame):
+            The real data for the desired column pair.
+        synthetic_column (pandas.Dataframe):
+            The synthetic data for the desired column pair.
+
+    Returns:
+        plotly.graph_objects._figure.Figure
+    """
+    columns = real_data.columns
+    real_data = real_data.copy()
+    real_data['Data'] = 'Real'
+    synthetic_data = synthetic_data.copy()
+    synthetic_data['Data'] = 'Synthetic'
+    all_data = pd.concat([real_data, synthetic_data], axis=0, ignore_index=True)
+
+    fig = px.scatter(
+        all_data,
+        x=columns[0],
+        y=columns[1],
+        color='Data',
+        color_discrete_map={'Real': DATACEBO_DARK, 'Synthetic': DATACEBO_LIGHT},
+        symbol='Data'
+    )
+
+    fig.update_layout(
+        title=f"Real vs. Synthetic Data for columns '{columns[0]}' and '{columns[1]}'",
+        plot_bgcolor=BACKGROUND_COLOR,
+    )
+
+    return fig
+
+
+def make_discrete_column_pair_plot(real_data, synthetic_data):
+    """Make a column pair plot for discrete data.
+
+    Args:
+        real_data (pandas.DataFrame):
+            The real data for the desired column pair.
+        synthetic_column (pandas.Dataframe):
+            The synthetic data for the desired column pair.
+
+    Returns:
+        plotly.graph_objects._figure.Figure
+    """
+    columns = real_data.columns
+    real_data = real_data.copy()
+    real_data['Data'] = 'Real'
+    synthetic_data = synthetic_data.copy()
+    synthetic_data['Data'] = 'Synthetic'
+    all_data = pd.concat([real_data, synthetic_data], axis=0, ignore_index=True)
+
+    fig = px.density_heatmap(
+        all_data,
+        x=columns[0],
+        y=columns[1],
+        facet_col='Data',
+        histnorm='probability'
+    )
+
+    fig.update_layout(
+        title_text=f"Real vs Synthetic Data for columns '{columns[0]}' and '{columns[1]}'",
+        coloraxis={'colorscale': [DATACEBO_DARK, DATACEBO_LIGHT]},
+    )
+
+    fig.for_each_annotation(lambda a: a.update(text=a.text.split('=')[-1] + ' Data'))
 
     return fig
 
 
+def make_mixed_column_pair_plot(real_data, synthetic_data):
+    """Make a column pair plot for mixed discrete and continuous column data.
+
+    Args:
+        real_data (pandas.DataFrame):
+            The real data for the desired column pair.
+        synthetic_column (pandas.Dataframe):
+            The synthetic data for the desired column pair.
+
+    Returns:
+        plotly.graph_objects._figure.Figure
+    """
+    columns = real_data.columns
+    real_data = real_data.copy()
+    real_data['Data'] = 'Real'
+    synthetic_data = synthetic_data.copy()
+    synthetic_data['Data'] = 'Synthetic'
+    all_data = pd.concat([real_data, synthetic_data], axis=0, ignore_index=True)
+
+    fig = px.box(
+        all_data,
+        x=columns[0],
+        y=columns[1],
+        color='Data',
+        color_discrete_map={'Real': DATACEBO_DARK, 'Synthetic': DATACEBO_LIGHT},
+    )
+
+    fig.update_layout(
+        title=f"Real vs. Synthetic Data for columns '{columns[0]}' and '{columns[1]}'",
+        plot_bgcolor=BACKGROUND_COLOR
+    )
+
+    return fig
+
+
+def get_column_pair_plot(real_data, synthetic_data, sdtypes):
+    """Return a plot of the real and synthetic data for a given column pair.
+
+    Args:
+        real_data (pandas.DataFrame):
+            The real data for the desired column pair.
+        synthetic_column (pandas.Dataframe):
+            The synthetic data for the desired column pair.
+        sdtypes (list[string]):
+            The data type of the column pair. The data type string must be one of
+            ('numerical', 'datetime', 'categorical', or 'boolean').
+
+    Returns:
+        plotly.graph_objects._figure.Figure
+    """
+    all_sdtypes = CONTINUOUS_SDTYPES + DISCRETE_SDTYPES
+    if not sdtypes[0] in all_sdtypes:
+        raise ValueError(f'sdtype of {sdtypes[0]} not recognized.')
+    if not sdtypes[1] in all_sdtypes:
+        raise ValueError(f'sdtype of {sdtypes[1]} not recognized.')
+
+    if all([t in DISCRETE_SDTYPES for t in sdtypes]):
+        return make_discrete_column_pair_plot(real_data, synthetic_data)
+
+    if sdtypes[0] == 'datetime':
+        real_data.iloc[:, 0] = convert_to_datetime(real_data.iloc[:, 0])
+    if sdtypes[1] == 'datetime':
+        real_data.iloc[:, 1] = convert_to_datetime(real_data.iloc[:, 1])
+
+    if all([t in CONTINUOUS_SDTYPES for t in sdtypes]):
+        return make_continuous_column_pair_plot(real_data, synthetic_data)
+    else:
+        return make_mixed_column_pair_plot(real_data, synthetic_data)
+
+
 def discretize_table_data(real_data, synthetic_data, metadata):
     """Create a copy of the real and synthetic data with discretized data.