def test_evaluate_tables_from_demo():
tables = load_demo(metadata=False)
new_meta = Metadata()
new_meta.add_table('users', data=tables['users'], primary_key='user_id')
new_meta.add_table('sessions',
data=tables['sessions'],
primary_key='session_id',
parent='users',
foreign_key='user_id')
transactions_fields = {
'timestamp': {
'type': 'datetime',
'format': '%Y-%m-%d'
}
}
new_meta.add_table('transactions',
tables['transactions'],
fields_metadata=transactions_fields,
primary_key='transaction_id',
parent='sessions')
sdv = SDV()
sdv.fit(new_meta, tables=tables)
sampled = sdv.sample_all()
table_scores = dict()
for table in new_meta.get_tables():
table_scores[table] = evaluate(sampled[table],
real=tables[table],
metadata=new_meta,
table_name=table)
evaluate(sampled, real=tables, metadata=new_meta)
def test_build_demo_metadata_from_tables():
"""Build metadata from the demo tables.
Then compare the built metadata with the demo one
to make sure that they are the same.
"""
tables = load_demo(metadata=False)
new_meta = Metadata()
new_meta.add_table('users', data=tables['users'], primary_key='user_id')
new_meta.add_table('sessions',
data=tables['sessions'],
primary_key='session_id',
parent='users',
foreign_key='user_id')
transactions_fields = {
'timestamp': {
'type': 'datetime',
'format': '%Y-%m-%dT%H:%M'
}
}
new_meta.add_table('transactions',
tables['transactions'],
fields_metadata=transactions_fields,
primary_key='transaction_id',
parent='sessions')
assert DEMO_METADATA == new_meta.to_dict()
def test_sdv():
metadata, tables = load_demo(metadata=True)
sdv = SDV()
sdv.fit(metadata, tables)
# Sample all
sampled = sdv.sample()
assert set(sampled.keys()) == {'users', 'sessions', 'transactions'}
assert len(sampled['users']) == 10
# Sample with children
sampled = sdv.sample('users', reset_primary_keys=True)
assert set(sampled.keys()) == {'users', 'sessions', 'transactions'}
assert len(sampled['users']) == 10
# Sample without children
users = sdv.sample('users', sample_children=False)
assert users.shape == tables['users'].shape
assert set(users.columns) == set(tables['users'].columns)
sessions = sdv.sample('sessions', sample_children=False)
assert sessions.shape == tables['sessions'].shape
assert set(sessions.columns) == set(tables['sessions'].columns)
transactions = sdv.sample('transactions', sample_children=False)
assert transactions.shape == tables['transactions'].shape
assert set(transactions.columns) == set(tables['transactions'].columns)
def test_sdv_multiparent():
metadata, tables = load_demo('got_families', metadata=True)
sdv = SDV()
sdv.fit(metadata, tables)
# Sample all
sampled = sdv.sample()
assert set(sampled.keys()) == {'characters', 'families', 'character_families'}
assert len(sampled['characters']) == 7
# Sample with children
sampled = sdv.sample('characters', reset_primary_keys=True)
assert set(sampled.keys()) == {'characters', 'character_families'}
assert len(sampled['characters']) == 7
assert 'family_id' in sampled['character_families']
# Sample without children
characters = sdv.sample('characters', sample_children=False)
assert characters.shape == tables['characters'].shape
assert set(characters.columns) == set(tables['characters'].columns)
families = sdv.sample('families', sample_children=False)
assert families.shape == tables['families'].shape
assert set(families.columns) == set(tables['families'].columns)
character_families = sdv.sample('character_families', sample_children=False)
assert character_families.shape == tables['character_families'].shape
assert set(character_families.columns) == set(tables['character_families'].columns)
def test_integration(self):
metadata, tables = load_demo(metadata=True)
sdv = SDV()
sdv.fit(metadata, tables)
synthetic = sdv.sample_all(20)
metrics = evaluate(metadata, tables, synthetic)
metrics.overall()
metrics.details()
metrics.highlights()
def test_integer_categoricals():
"""Ensure integer categoricals are still sampled as integers.
The origin of this tests can be found in the github issue #194:
https://github.com/sdv-dev/SDV/issues/194
"""
metadata, tables = load_demo(metadata=True)
metadata_dict = metadata.to_dict()
metadata_dict['tables']['users']['fields']['age'] = {'type': 'categorical'}
sdv = SDV()
sdv.fit(metadata, tables)
sampled = sdv.sample()
for name, table in tables.items():
assert (sampled[name].dtypes == table.dtypes).all()
"""
Running the SDV basic tutorial using their example dataset.
"""
from sdv import load_demo
from sdv import SDV
# Grab the demo data
metadata, tables = load_demo(metadata=True)
print(metadata)
# Run the basic fit
sdv = SDV()
sdv.fit(metadata, tables)
print("done fit")
sdv.save('sdv.pkl')
