def test_functional_compounder_double_aggregation(dataset):
compounder = FunctionalCompounder()
groups = ['type', 'year']
aggregations = ['sum:price', 'count:name']
aggregator = compounder.compound(groups=groups, aggregations=aggregations)
result = aggregator(dataset)
assert result == ([{
'type': 'sedan',
'year': '2018',
'sum_price': 21000,
'count_name': 1
}, {
'type': 'sedan',
'year': '2019',
'sum_price': 90000,
'count_name': 2
}, {
'type': 'sedan',
'year': '2020',
'sum_price': 121000,
'count_name': 2
}, {
'type': 'suv',
'year': '2019',
'sum_price': 67000,
'count_name': 1
}, {
'type': 'suv',
'year': '2020',
'sum_price': 47000,
'count_name': 1
}, {
'type': 'truck',
'year': '2019',
'sum_price': 130000,
'count_name': 2
}, {
'type': 'truck',
'year': '2020',
'sum_price': 43000,
'count_name': 1
}, {
'type': 'van',
'year': '2018',
'sum_price': 45000,
'count_name': 1
}, {
'type': 'van',
'year': '2020',
'sum_price': 30000,
'count_name': 1
}])
def test_functional_compounder_two_groups_inverted(dataset):
compounder = FunctionalCompounder()
groups = ['type', 'year']
aggregator = compounder.compound(groups=groups)
result = aggregator(dataset)
assert result == [{
'type': 'sedan',
'year': '2018',
'count_type': 1
}, {
'type': 'sedan',
'year': '2019',
'count_type': 2
}, {
'type': 'sedan',
'year': '2020',
'count_type': 2
}, {
'type': 'suv',
'year': '2019',
'count_type': 1
}, {
'type': 'suv',
'year': '2020',
'count_type': 1
}, {
'type': 'truck',
'year': '2019',
'count_type': 2
}, {
'type': 'truck',
'year': '2020',
'count_type': 1
}, {
'type': 'van',
'year': '2018',
'count_type': 1
}, {
'type': 'van',
'year': '2020',
'count_type': 1
}]
def test_functional_compounder_other_key(dataset):
compounder = FunctionalCompounder()
groups = ['year']
aggregator = compounder.compound(groups=groups)
result = aggregator(dataset)
assert result == [{
'year': '2018',
'count_year': 2
}, {
'year': '2019',
'count_year': 5
}, {
'year': '2020',
'count_year': 5
}]
def test_functional_compounder_composite_functions(dataset):
compounder = FunctionalCompounder()
test_tuples = [
('avg', 'price', 49500),
('count', 'price', 12),
('max', 'price', 70000),
('median', 'price', 49000.0),
('min', 'price', 21000),
('stdev', 'price', 17479.858539057306),
('sum', 'price', 594000),
('var', 'price', 305545454.54545456),
]
for test_tuple in test_tuples:
result = compounder._build_composite(test_tuple[0], test_tuple[1],
dataset)
assert result == test_tuple[2]
def test_functional_compounder_single_aggregation(dataset):
compounder = FunctionalCompounder()
groups = ['year']
aggregations = ['sum:price']
aggregator = compounder.compound(groups=groups, aggregations=aggregations)
result = aggregator(dataset)
assert result == [{
'year': '2018',
'sum_price': 66000
}, {
'year': '2019',
'sum_price': 287000
}, {
'year': '2020',
'sum_price': 241000
}]
def test_functional_compounder_compound(dataset):
compounder = FunctionalCompounder()
groups = ['type']
aggregator = compounder.compound(groups=groups)
result = aggregator(dataset)
assert result == [{
'type': 'sedan',
'count_type': 5
}, {
'type': 'suv',
'count_type': 2
}, {
'type': 'truck',
'count_type': 3
}, {
'type': 'van',
'count_type': 2
}]
def test_functional_compounder_instantiation():
compounder = FunctionalCompounder()
assert compounder is not None