Unable visualize just the real data (or just the synthetic data) in a multi-table setting

[npatki]

Environment Details

• SDV version: 1.15.0
• Python version: 3.10
• Operating System: Linux (Google Colab)

Error Description

The latest update of SDMetrics allows you to plot just the real data (or just the synthetic data) by supplying the other dataset as None. This is useful for explorations an investigations.

In SDV, we have a wrapper around SDMetrics. I expect that these visualizations should have the same set of features as SDMetrics. However, I see that the ability to set None doesn't work -- it produces an error instead.

Steps to reproduce

from sdv.datasets.demo import download_demo
from sdv.evaluation.multi_table import get_column_plot

data, metadata = download_demo(
    modality='multi_table',
    dataset_name='fake_hotels'
)

fig = get_column_plot(
    real_data=data,
    synthetic_data=None, # I should be able set either this or the real_data to None
    metadata=metadata,
    table_name='guests',
    column_name='amenities_fee'
)
    
fig.show()

[/usr/local/lib/python3.10/dist-packages/sdv/evaluation/multi_table.py](https://localhost:8080/#) in get_column_plot(real_data, synthetic_data, metadata, table_name, column_name, plot_type)
     79     metadata = metadata.tables[table_name]
     80     real_data = real_data[table_name]
---> 81     synthetic_data = synthetic_data[table_name]
     82     return single_table_visualization.get_column_plot(
     83         real_data,

TypeError: 'NoneType' object is not subscriptable

Additional Context

Note that this works correctly in the single-table context. Only multi-table is erroring.

# this works as expected!
from sdv.datasets.demo import download_demo
from sdv.evaluation.single_table import get_column_plot

data, metadata = download_demo(
    modality='single_table',
    dataset_name='fake_hotel_guests'
)


fig = get_column_plot(
    real_data=data,
    synthetic_data=None,
    metadata=metadata,
    column_name='amenities_fee'
)
    
fig.show()