def test_overlap_but_on_different_day():
df = pd.DataFrame({
'date': ['2020-08-10', '2020-08-10', '2020-08-11', '2020-08-11'],
'song': ['S1', 'S2', 'S1', 'S3'],
'artist': ['A1', 'A1', 'A1', 'A2'],
'radio': ['rfm', 'rfm', 'comercial', 'comercial'],
'time': ['00:01', '00:04', '00:02', '00:05']
})
df = _add_datetime_to_df(df)
result = rra.overlaps(df)
assert len(result) == 0
def test_overlap_consecutive_days():
df = pd.DataFrame({
'date': ['2020-08-10', '2020-08-11', '2020-08-11', '2020-08-11'],
'song': ['S1', 'S2', 'S1', 'S3'],
'artist': ['A1', 'A1', 'A1', 'A2'],
'radio': ['rfm', 'rfm', 'comercial', 'comercial'],
'time': ['23:58', '00:03', '00:02', '00:07']
})
df = _add_datetime_to_df(df)
result = rra.overlaps(df)
assert len(result) == 1
result = result.iloc[0]
assert result['song'] == 'S1' and result['artist'] == 'A1'
assert len(result['radios'].split(' | ')) == 2
assert 'rfm' in result['radios'] and 'comercial' in result['radios']
def test_three_overlaps():
df = pd.DataFrame({
'date': [
'2020-08-10', '2020-08-10', '2020-08-10', '2020-08-10',
'2020-08-10', '2020-08-10'
],
'song': ['S1', 'S2', 'S1', 'S3', 'S1', 'S5'],
'artist': ['A1', 'A1', 'A1', 'A2', 'A1', 'A5'],
'radio': ['rfm', 'rfm', 'comercial', 'comercial', 'megafm', 'megafm'],
'time': ['15:38', '15:42', '15:39', '15:43', '15:41', '15:45']
})
df = _add_datetime_to_df(df)
result = rra.overlaps(df)
assert len(result) == 1
result = result.iloc[0]
assert result['song'] == 'S1' and result['artist'] == 'A1'
assert len(result['radios'].split(' | ')) == 3
assert sorted(
result['radios'].split(' | ')) == ['comercial', 'megafm', 'rfm']
def test_multiple_overlaps():
"""
Tests multiple overlaps, with unordered times, and a complex three-overlap.
The three-overlap takes the form:
[ ]
[ ]
[ ]
That is, the first and last plays do not overlap directly, only through the middle play.
"""
df = pd.DataFrame({
'date': [
'2020-08-10', '2020-08-10', '2020-08-10', '2020-08-10',
'2020-08-10', '2020-08-10', '2020-08-10', '2020-08-10',
'2020-08-10'
],
'song': ['S1', 'S2', 'S3', 'S4', 'S2', 'S5', 'S3', 'S2', 'S5'],
'artist': ['A1', 'A2', 'A3', 'A4', 'A2', 'A5', 'A3', 'A2', 'A5'],
'radio': [
'rfm', 'rfm', 'rfm', 'comercial', 'comercial', 'comercial',
'megafm', 'megafm', 'megafm'
],
'time': [
'15:38', '15:42', '15:46', '15:37', '15:41', '15:45', '15:41',
'15:37', '15:35'
]
})
df = _add_datetime_to_df(df)
result = rra.overlaps(df)
assert len(result) == 2
first = result.iloc[0]
assert first['song'] == 'S2' and first['artist'] == 'A2'
assert len(first['radios'].split(' | ')) == 3
assert sorted(
first['radios'].split(' | ')) == ['comercial', 'megafm', 'rfm']
second = result.iloc[1]
assert second['song'] == 'S3' and second['artist'] == 'A3'
assert len(second['radios'].split(' | ')) == 2
assert sorted(second['radios'].split(' | ')) == ['megafm', 'rfm']