def test_multiplots():
hotel_stays = samples.hotel_stays_sample()
fig, ax = plt.subplots(nrows=4,figsize=(36,8))
hcl = hotel_stays.clip(lbound=pd.Timestamp(2016,8,1))
hcu = hotel_stays.clip(ubound=pd.Timestamp(2016,5,1))
hcl.plot(ax=ax[0],color='g')
hcu.plot(ax=ax[0],color='r')
hcl.smooth_plot(ax=ax[0],color='black',smooth_factor=10**5)
hcu.smooth_plot(ax=ax[0],color='black',smooth_factor=10**5)
hotel_stays.plot(ax=ax[1],color='r')
hotel_stays.smooth_plot(ax=ax[1],color='g',linewidth=5,smooth_factor=10**5)
step_clip_start = Steps(start=pd.Timestamp(2016,6,1))
step_clip_end = Steps(end=pd.Timestamp(2016,6,1))
(hotel_stays*step_clip_start).plot(ax=ax[2])
(hotel_stays*step_clip_end).plot(ax=ax[3])
perform_plot_comparison('hotel_multiplots', fig)
def test_temperature_study():
df_temps = pd.read_csv(
r'https://raw.githubusercontent.com/TangleSpace/hotstepper-data/master/data/daily-min-temperatures.csv',
parse_dates=['Date'])
temps_steps = Steps.read_dataframe(df_temps,
start='Date',
weight='Temp',
convert_delta=True)
gt20_df = df_temps[
df_temps.Temp > 20]['Temp'].count() / df_temps['Temp'].count()
temps_step20 = temps_steps > 20
gt20 = (temps_step20.clamp(lbound=temps_step20.first() -
pd.Timedelta(days=1)).normalise().integrate() /
temps_steps.normalise().integrate())
np.testing.assert_almost_equal(gt20, gt20_df)
#pacf
x, y = temps_steps.pacf(50)
np.testing.assert_almost_equal(y, temperature_pacf50)
np.testing.assert_almost_equal(x, list(range(51)))
#acf
x, y = temps_steps.acf(50)
np.testing.assert_almost_equal(y, temperature_acf50)
np.testing.assert_almost_equal(x, list(range(51)))
#histogram, 10 bins
x, y = temps_steps.histogram(bins=10)
np.testing.assert_almost_equal(y, temperature_hist10y)
np.testing.assert_almost_equal(x, temperature_hist10x)
def daily_temperature_sample():
df = pd.read_csv(os.path.join(DATA_DIR, 'daily-min-temperatures.csv'),
parse_dates=['Date'])
return Steps.read_dataframe(df,
start='Date',
weight='Temp',
convert_delta=True)
def test_vessel_multiplots():
fig,ax = plt.subplots(nrows=5,figsize=(20,12))
vsteps = samples.vessel_queue_sample()
v_ul_clip = vsteps.clip(lbound=pd.Timestamp(2020,2,1),ubound=pd.Timestamp(2020,2,10))
v_ul_clip.plot(ax=ax[1])
(v_ul_clip>=8).plot(ax=ax[1],color='blue')
((v_ul_clip>=8)/v_ul_clip).plot(ax=ax[1],color='red')
(v_ul_clip>=8).normalise().plot(ax=ax[1],color='black')
vsteps.clip(lbound=pd.Timestamp(2020,2,1),ubound=pd.Timestamp(2020,5,1)).plot(ax=ax[0],color='r')
clip_step_end = Steps(end=pd.Timestamp(2020,1,10))
steps_end = Steps(True).add_direct([None],[pd.Timestamp(2020,1,10)])
(vsteps*clip_step_end).plot(ax=ax[2],color='black')
(vsteps*steps_end).plot(ax=ax[2],color='r')
vsteps.clip(ubound=pd.Timestamp(2020,1,10)).plot(ax=ax[2],color='g')
vsteps.plot(ax=ax[4])
vsteps.smooth_plot(ax=ax[4],color='g',linewidth=3)
vc = vsteps.clip(ubound=pd.Timestamp(2020,1,10))
vc.plot(ax=ax[3])
vc7 = vc>9
vc7.plot(ax=ax[3],color='g')
vc7.invert(3).plot(ax=ax[3],color='black')
perform_plot_comparison('vessel_multiplots', fig)
def vessel_queue_sample():
df = pd.read_csv(os.path.join(DATA_DIR, 'vessel_queue.csv'),
parse_dates=['enter', 'leave'],
dayfirst=True)
return Steps.read_dataframe(df, 'enter', 'leave', use_datetime=True)
def hotel_stays_sample():
df = pd.read_csv(os.path.join(DATA_DIR, 'hotel_stays.csv'),
parse_dates=['check_in', 'check_out'],
dayfirst=True)
return Steps.read_dataframe(df, 'check_in', 'check_out', use_datetime=True)
def page_view_sample():
df = pd.read_csv(os.path.join(DATA_DIR, 'page_views.csv'))
return Steps.read_dataframe(df, 'start', 'ends')