def test_normalize_dataframe():
dic = {
'team': [
'Red', 'Red', 'Red', 'Orange', 'Orange', 'Yellow', 'Yellow',
'Green', 'Green', 'Blue'
],
'jersey_num': [1, 2, 3, 1, 2, 1, 5, 8, 2, 2],
'player_name': ['A', 'B', 'C', 'D', 'A', 'E', 'B', 'A', 'G', 'H'],
'city': [
'boston', 'boston', 'boston', 'chicago', 'chicago', 'honolulu',
'honolulu', 'boston', 'boston', 'austin'
],
'state': ['MA', 'MA', 'MA', 'IL', 'IL', 'HI', 'HI', 'MA', 'MA', 'TX']
}
df = pd.DataFrame(dic)
deps = classes.Dependencies(
{
'team': [['player_name', 'jersey_num']],
'jersey_num': [['player_name', 'team']],
'player_name': [['team', 'jersey_num']],
'city': [['team'], ['state'], ['player_name', 'jersey_num']],
'state': [['team'], ['player_name', 'jersey_num'], ['city']]
}, ['team', 'jersey_num'])
depdf = normalize.DepDF(deps, df, deps.get_prim_key())
normalize.normalize_dataframe(depdf)
new_dfs = depdf.return_dfs()
assert len(new_dfs) == 3
dic_one = {
'team': [
'Red', 'Red', 'Red', 'Orange', 'Orange', 'Yellow', 'Yellow',
'Green', 'Green', 'Blue'
],
'jersey_num': [1, 2, 3, 1, 2, 1, 5, 8, 2, 2],
'player_name': ['A', 'B', 'C', 'D', 'A', 'E', 'B', 'A', 'G', 'H']
}
dic_two = {
'team': ['Red', 'Orange', 'Yellow', 'Green', 'Blue', 'Blue'],
'city':
['boston', 'chicago', 'honolulu', 'boston', 'austin', 'austin']
}
dic_three = {
'city': ['boston', 'chicago', 'honolulu', 'austin', 'austin'],
'state': ['MA', 'IL', 'HI', 'TX', 'TX']
}
assert new_dfs[0].equals(
normalize.drop_primary_dups(pd.DataFrame(dic_one),
['team', 'jersey_num']))
assert new_dfs[1].equals(
normalize.drop_primary_dups(pd.DataFrame(dic_two), ['team']))
assert new_dfs[2].equals(
normalize.drop_primary_dups(pd.DataFrame(dic_three), ['city']))
def test_make_indexes():
dic = {
"id": [0, 1, 2, 3, 4, 5, 6, 7, 8, 9],
"month":
['dec', 'dec', 'jul', 'jul', 'dec', 'jul', 'jul', 'jul', 'dec', 'jul'],
"hemisphere": ['N', 'N', 'N', 'N', 'S', 'S', 'S', 'S', 'S', 'N'],
"is_winter":
[True, True, False, False, False, True, True, True, False, False]
}
df = pd.DataFrame(dic)
deps = classes.Dependencies(
{
'id': [],
'month': [['id'], ['hemisphere', 'is_winter']],
'hemisphere': [['month', 'is_winter'], ['id']],
'is_winter': [['month', 'hemisphere'], ['id']]
}, ['id'])
depdf = normalize.DepDF(deps, df, deps.get_prim_key())
normalize.normalize_dataframe(depdf)
normalize.make_indexes(depdf)
new_dfs = depdf.return_dfs()
mask = (new_dfs[1]['month'] == 'dec') & (new_dfs[1]['hemisphere'] == 'N')
val = new_dfs[1][mask][new_dfs[1].columns[0]].iloc[0]
assert new_dfs[0][new_dfs[1].columns[0]][0] == val
assert new_dfs[0][new_dfs[1].columns[0]][1] == val
mask = (new_dfs[1]['month'] == 'jul') & (new_dfs[1]['hemisphere'] == 'N')
val = new_dfs[1][mask][new_dfs[1].columns[0]].iloc[0]
assert new_dfs[0][new_dfs[1].columns[0]][2] == val
assert new_dfs[0][new_dfs[1].columns[0]][3] == val
assert new_dfs[0][new_dfs[1].columns[0]][9] == val
mask = (new_dfs[1]['month'] == 'dec') & (new_dfs[1]['hemisphere'] == 'S')
val = new_dfs[1][mask][new_dfs[1].columns[0]].iloc[0]
assert new_dfs[0][new_dfs[1].columns[0]][4] == val
assert new_dfs[0][new_dfs[1].columns[0]][8] == val
mask = (new_dfs[1]['month'] == 'jul') & (new_dfs[1]['hemisphere'] == 'S')
val = new_dfs[1][mask][new_dfs[1].columns[0]].iloc[0]
assert new_dfs[0][new_dfs[1].columns[0]][5] == val
assert new_dfs[0][new_dfs[1].columns[0]][6] == val
assert new_dfs[0][new_dfs[1].columns[0]][7] == val
# Make sure new column names are sorted
assert 'hemisphere_month' in new_dfs[0].columns
assert 'hemisphere_month' in new_dfs[1].columns