def test_analysis_with_empty_df(self):
"Empty dataframes should not raise errors"
anls = analysis.DailyAgg(df=pd.DataFrame(), agg='max')
self.assertIsInstance(anls.result, pd.DataFrame)
js = anls.to_json()
self.assertIsInstance(pd.read_json(js), pd.DataFrame)
def test_init_DailyAgg(self):
"Initialize a DailyAgg, check attributes of result"
index = pd.DatetimeIndex(start='20160101 00:51:15',
freq='h',
periods=80)
df = pd.DataFrame(index=index, data=np.arange(0, 80, 1), columns=['A'])
anls = analysis.DailyAgg(df, agg='min')
self.assertTrue(hasattr(anls, 'result'))
def test_json(self):
"Verify the to_json method"
index = pd.DatetimeIndex(start='20160101 00:51:15',
freq='h',
periods=80)
df = pd.DataFrame(index=index,
data={
'A': np.arange(0, 80, 1),
'B': np.zeros(80)
})
js = analysis.DailyAgg(df,
agg='min',
starttime=dt.time(hour=3, minute=34),
endtime=dt.time(hour=22, minute=55)).to_json()
self.assertIsInstance(pd.read_json(js), pd.DataFrame)
def test_DailyAgg_analysis(self):
"Verify result of analyses"
index = pd.DatetimeIndex(start='20160101 00:51:15',
freq='h',
periods=80)
df = pd.DataFrame(index=index,
data={
'A': np.arange(0, 80, 1),
'B': np.zeros(80)
})
anls = analysis.DailyAgg(df, agg='min')
dayindex = pd.DatetimeIndex(start='20160101', freq='D', periods=4)
self.assertTrue((anls.result.index == dayindex).all())
self.assertTrue((anls.result['A'] == pd.Series(data=[0, 24, 48, 72],
index=dayindex)).all())
def test_DailyAgg_analysis_with_timelimits(self):
"Check results when limiting the hours for the analysis"
index = pd.DatetimeIndex(start='20160101 00:51:15',
freq='h',
periods=80)
df = pd.DataFrame(index=index,
data={
'A': np.arange(0, 80, 1),
'B': np.zeros(80)
})
anls = analysis.DailyAgg(df,
agg='min',
starttime=dt.time(hour=3, minute=34),
endtime=dt.time(hour=22, minute=55))
dayindex = pd.DatetimeIndex(start='20160101', freq='D', periods=4)
self.assertTrue((anls.result.index == dayindex).all())
self.assertTrue((anls.result['A'] == pd.Series(data=[3, 27, 51, 75],
index=dayindex)).all())
def test_run_analysis_twice_different_args(self):
"Running an analysis twice should give same result"
index = pd.DatetimeIndex(start='20160101 00:51:15',
freq='h',
periods=80)
df = pd.DataFrame(index=index,
data={
'A': np.arange(0, 80, 1),
'B': np.zeros(80)
})
anls = analysis.DailyAgg(df,
agg='min',
starttime=dt.time(hour=3, minute=34),
endtime=dt.time(hour=22, minute=55))
result1 = anls.result.copy()
anls.do_analysis(agg='max',
starttime=dt.time(hour=3, minute=34),
endtime=dt.time(hour=22, minute=55))
result2 = anls.result.copy()
self.assertFalse((result1 == result2).all().all())