def test_load_campbell_scientific():
data = bw.load_campbell_scientific(
'../datasets/demo/campbell_scientific_demo_data.csv')
data1 = bw.load_campbell_scientific(
'../datasets/demo/campbell_scientific_demo_data1.csv', dayfirst=True)
assert (data['2016-01-09 15:30:00':'2016-01-10 23:50:00'].fillna(-999) ==
data1['2016-01-09 15:30:00':'2016-01-10 23:50:00'].fillna(-999)
).all().all()
def test_dist():
df = bw.load_campbell_scientific(bw.datasets.demo_campbell_scientific_data)
dist = bw.dist(df[['Spd40mN']],
bins=[0, 8, 12, 21],
bin_labels=['normal', 'gale', 'storm'])
# For distribution of %frequency of wind speeds
dist = bw.dist(df.Spd40mN,
bins=[0, 8, 12, 21],
bin_labels=['normal', 'gale', 'storm'])
# For distribution of mean temperature
temp_dist = bw.dist(df.T2m,
bins=[-10, 4, 12, 18, 30],
aggregation_method='mean')
# For custom aggregation function
def custom_agg(x):
return x.mean() + (2 * x.std())
temp_dist = bw.dist(df.T2m,
bins=[-10, 4, 12, 18, 30],
aggregation_method=custom_agg)
# For distribution of mean wind speeds with respect to temperature
spd_dist = bw.dist(df.Spd40mN,
var_to_bin_against=df.T2m,
bins=[-10, 4, 12, 18, 30],
bin_labels=['freezing', 'cold', 'mild', 'hot'],
aggregation_method='mean')
def test_apply_cleaning():
data = bw.load_campbell_scientific(
bw.datasets.demo_campbell_scientific_data)
data_clnd2 = bw.apply_cleaning_windographer(
data, bw.datasets.demo_windographer_flagging_log2, dayfirst=True)
data_clnd3 = bw.apply_cleaning(data, bw.datasets.demo_cleaning_file)
data_clnd4 = bw.apply_cleaning(data,
'../datasets/demo/demo_cleaning_file2.csv',
dayfirst=True)
data_clnd5 = bw.apply_cleaning(data,
'../datasets/demo/demo_cleaning_file3.csv',
dayfirst=True)
assert (data_clnd2.drop(['RECORD', 'Site', 'LoggerID'],
axis=1).fillna(-999) == data_clnd3.drop(
['RECORD', 'Site', 'LoggerID'],
axis=1).fillna(-999)).all().all()
assert (data_clnd3.drop(['RECORD', 'Site', 'LoggerID'],
axis=1).fillna(-999) == data_clnd4.drop(
['RECORD', 'Site', 'LoggerID'],
axis=1).fillna(-999)).all().all()
assert (data_clnd3.drop(['RECORD', 'Site', 'LoggerID'],
axis=1).fillna(-999) == data_clnd5.drop(
['RECORD', 'Site', 'LoggerID'],
axis=1).fillna(-999)).all().all()
def test_freq_table():
df = bw.load_campbell_scientific(bw.datasets.demo_campbell_scientific_site_data)
graph = bw.freq_table(df[['Spd40mN']], df[['Dir38mS']])
graph, tab = bw.freq_table(df.Spd40mN, df.Dir38mS, return_data=True)
# Calling with user defined dir_bin labels BUGGY
graph, tab = bw.freq_table(df.Spd40mN, df.Dir38mS, direction_bin_array=[0, 90, 160, 210, 360],
direction_bin_labels=['lowest', 'lower', 'mid', 'high'], return_data=True)
assert (tab.columns == ['lowest','lower','mid','high']).all()
tab = bw.freq_table(df.Spd40mN, df.Dir38mS, plot_bins=[0, 3, 6, 9, 12, 15, 41],
plot_labels=['0-3 m/s', '4-6 m/s', '7-9 m/s', '10-12 m/s', '13-15 m/s', '15+ m/s'],
return_data=True)
#Calling with user defined var_bin labels
graph, tab = bw.freq_table(df.Spd40mN, df.Dir38mS, var_bin_array=[0, 10, 15, 50],
var_bin_labels=['low', 'mid', 'high'], plot_bins=[0, 10, 15, 50], plot_labels=None,
return_data=True)
tab = bw.freq_table(df.Spd40mN, df.Dir38mS, var_bin_array=[0, 8, 14, 41], var_bin_labels=['low', 'mid', 'high'],
direction_bin_array=[0, 90, 130, 200, 360],
direction_bin_labels=['northerly', 'easterly', 'southerly', 'westerly'],
plot_bins=[0, 8, 14, 41], plot_labels=None, return_data=True)
# var_bin_labels=['operating','shutdow','dangerous'],
temp_rose, temp_freq_tab = bw.freq_table(df.T2m, df.Dir78mS, var_bin_array=[-10, 0, 10, 20],
var_bin_labels=['low', 'mid', 'high'],
plot_bins=[-10, 0, 10, 20], plot_labels=None,
return_data=True)
def test_export_to_csv():
df = bw.load_campbell_scientific(bw.datasets.demo_campbell_scientific_site_data)
bw.export_csv(df, file_name='export_to_csv')
bw.export_csv(df, file_name='export_to_csv_tab.tab', sep='\t')
def test_average_data_by_period():
data = bw.load_campbell_scientific(bw.datasets.demo_campbell_scientific_site_data)
#hourly averages
bw.average_data_by_period(data.Spd80mN, period='1H')
#hourly average with coverage filtering
bw.average_data_by_period(data.Spd80mN, period='1H', coverage_threshold=0.9)
bw.average_data_by_period(data.Spd80mN, period='1H', coverage_threshold=1)
#return coverage with filtering
bw.average_data_by_period(data.Spd80mN, period='1H', coverage_threshold=0.9,
return_coverage=True)
# return coverage without filtering
bw.average_data_by_period(data.Spd80mN, period='1H', return_coverage=True)
#monthly averages
bw.average_data_by_period(data.Spd80mN, period='1M')
#hourly average with coverage filtering
bw.average_data_by_period(data.Spd80mN, period='1M', coverage_threshold=0.9)
bw.average_data_by_period(data.Spd80mN, period='1M', coverage_threshold=1)
#return coverage with filtering
bw.average_data_by_period(data.Spd80mN, period='1M', coverage_threshold=0.9, return_coverage=True)
# return coverage without filtering
bw.average_data_by_period(data.Spd80mN, period='1M', return_coverage=True)
#weekly averages
bw.average_data_by_period(data.Spd80mN, period='2W')
#hourly average with coverage filtering
bw.average_data_by_period(data.Spd80mN, period='2W', coverage_threshold=0.9)
bw.average_data_by_period(data.Spd80mN, period='2W', coverage_threshold=1)
#return coverage with filtering
bw.average_data_by_period(data.Spd80mN, period='2W', coverage_threshold=0.9, return_coverage=True)
# return coverage without filtering
bw.average_data_by_period(data.Spd80mN, period='2W', return_coverage=True)
def test_average_data_by_period():
data = bw.load_campbell_scientific(
bw.datasets.demo_campbell_scientific_site_data)
bw.average_data_by_period(data[['Spd80mN']], period='1H')
# hourly averages
bw.average_data_by_period(data.Spd80mN, period='1H')
# hourly average with coverage filtering
bw.average_data_by_period(data.Spd80mN,
period='1H',
coverage_threshold=0.9)
bw.average_data_by_period(data.Spd80mN, period='1H', coverage_threshold=1)
# return coverage with filtering
bw.average_data_by_period(data.Spd80mN,
period='1H',
coverage_threshold=0.9,
return_coverage=True)
# return coverage without filtering
bw.average_data_by_period(data.Spd80mN, period='1H', return_coverage=True)
# monthly averages
bw.average_data_by_period(data.Spd80mN, period='1M')
# hourly average with coverage filtering
bw.average_data_by_period(data.Spd80mN,
period='1M',
coverage_threshold=0.9)
bw.average_data_by_period(data.Spd80mN, period='1M', coverage_threshold=1)
# return coverage with filtering
bw.average_data_by_period(data.Spd80mN,
period='1M',
coverage_threshold=0.9,
return_coverage=True)
# return coverage without filtering
bw.average_data_by_period(data.Spd80mN, period='1M', return_coverage=True)
# weekly averages
bw.average_data_by_period(data.Spd80mN, period='2W')
# hourly average with coverage filtering
bw.average_data_by_period(data.Spd80mN,
period='2W',
coverage_threshold=0.9)
bw.average_data_by_period(data.Spd80mN, period='2W', coverage_threshold=1)
# return coverage with filtering
bw.average_data_by_period(data.Spd80mN,
period='2W',
coverage_threshold=0.9,
return_coverage=True)
# return coverage without filtering
bw.average_data_by_period(data.Spd80mN, period='2W', return_coverage=True)
average_monthly_speed = bw.average_data_by_period(dummy_data_frame(), '1M')
average_annual_speed = bw.average_data_by_period(dummy_data_frame(), '1As')
# round annual wind speed
print(round(average_annual_speed, 2))
# test average wind speed for each month
for i in range(0, 11):
assert average_monthly_speed.iloc[i].item() == i + 1
# test average annual wind speed
assert round(average_annual_speed.iloc[0].item(), 1) == 6.5
def test_TI_twelve_by_24():
df = bw.load_campbell_scientific(bw.datasets.demo_campbell_scientific_site_data)
bw.TI.twelve_by_24(df[['Spd60mN']], df[['Spd60mNStd']])
bw.TI.twelve_by_24(df.Spd60mN, df.Spd60mNStd)
bw.TI.twelve_by_24(df.Spd60mN, df.Spd60mNStd, return_data=True)
bw.TI.twelve_by_24(df.Spd60mN, df.Spd60mNStd, return_data=True, var_name_label='Speed 60 m N m/s')
bw.TI.twelve_by_24(df.Spd60mN, df.Spd60mNStd, var_name_label='Speed 60 m N m/s')
bw.TI.twelve_by_24(df.Spd40mN, df.Spd40mNStd)
assert 1 == 1
def test_apply_cleaning_windographer():
data = bw.load_campbell_scientific(
bw.datasets.demo_campbell_scientific_data)
data_clnd = bw.apply_cleaning_windographer(
data, bw.datasets.demo_windographer_flagging_log)
data_clnd2 = bw.apply_cleaning_windographer(
data, bw.datasets.demo_windographer_flagging_log2, dayfirst=True)
assert (data_clnd2.fillna(-999) == data_clnd.fillna(-999)).all().all()
def test_plot_timeseries():
df = bw.load_campbell_scientific(
bw.datasets.demo_campbell_scientific_site_data)
graph = bw.plot_timeseries(df[['Spd40mN', 'Spd60mS', 'T2m']])
graph = bw.plot_timeseries(df.Spd40mN, date_from='2016-04-21')
graph = bw.plot_timeseries(df.Spd40mN,
date_from='2016-01-21',
date_to='2016-02-28')
assert True
def test_freq_table():
df = bw.load_campbell_scientific(
bw.datasets.demo_campbell_scientific_site_data)
temp_rose, temp_freq_tab = bw.freq_table(
df.T2m,
df.Dir78mS,
var_bin_array=[-10, 0, 10, 20],
var_bin_labels=['low', 'mid', 'high'],
return_data=True)
assert 1 == 1
def test_dist_matrix_by_direction_sector():
df = bw.load_campbell_scientific(bw.datasets.demo_campbell_scientific_site_data)
bw.dist_matrix_by_dir_sector(var_series=df.Spd80mN, var_to_bin_by_series=df.Spd80mN, direction_series=df.Dir78mS,
aggregation_method='count')
matrix = bw.dist_matrix_by_dir_sector(df.Spd40mN, df.T2m, df.Dir38mS,
var_to_bin_by_array=[-8, -5, 5, 10, 15, 20, 26],
direction_bin_array=[0, 90, 180, 270, 360],
direction_bin_labels=['north', 'east', 'south', 'west'])
matrix = bw.dist_matrix_by_dir_sector(df.Spd40mN, df.T2m, df.Dir38mS,
var_to_bin_by_array=[-8, -5, 5, 10, 15, 20, 26], sectors=8)
assert True
def test_coverage():
data = bw.load_campbell_scientific(bw.datasets.demo_campbell_scientific_site_data)
# hourly coverage
data_hourly = bw.coverage(data[['Spd80mN']], period='1H')
data_hourly = bw.coverage(data.Spd80mN, period='1H')
# monthly_coverage
data_monthly = bw.coverage(data.Spd80mN, period='1M')
# monthly_coverage of variance
data_monthly= bw.coverage(data.Spd80mN, period='1M', aggregation_method='var')
assert True
def test_dist_by_dir_sector():
df = bw.load_campbell_scientific(bw.datasets.demo_campbell_scientific_site_data)
rose = bw.dist_by_dir_sector(df[['Spd40mN']], df[['Dir38mS']])
rose = bw.dist_by_dir_sector(df.Spd40mN, df.Dir38mS)
rose, distribution = bw.dist_by_dir_sector(df.Spd40mN, df.Dir38mS,
direction_bin_array=[0, 90, 130, 200, 360],
direction_bin_labels=['northerly', 'easterly', 'southerly',
'westerly'],
return_data=True)
rose, distribution = bw.dist_by_dir_sector(df.Spd40mN, df.Dir38mS, aggregation_method='std', return_data=True)
def test_TI_by_speed():
df = bw.load_campbell_scientific(bw.datasets.demo_campbell_scientific_site_data)
TI_by_speed = bw.TI.by_speed(df[['Spd80mN']], df[['Spd80mNStd']])
TI_by_speed = bw.TI.by_speed(df.Spd80mN, df.Spd80mNStd)
#60 percentile
TI_by_speed_60 = bw.TI.by_speed(df.Spd80mN, df.Spd80mNStd, percentile=60, return_data=True)
#bin_array
TI_by_speed = bw.TI.by_speed(df.Spd80mN, df.Spd80mNStd, speed_bin_array=[0, 10, 14, 51],
speed_bin_labels=['low', 'mid', 'high'], return_data=True)
# assert TI_by_speed.index == ['low', 'mid', 'high']
assert True
def test_get_data_resolution():
import warnings
series1 = bw.load_campbell_scientific(bw.datasets.demo_campbell_scientific_site_data)['Spd80mS'].index
assert bw._get_data_resolution(series1).seconds == 600
series2 = pd.date_range('2010-01-01', periods=150, freq='H')
assert bw._get_data_resolution(series2).seconds == 3600
#hourly series with one instance where difference between adjacent timestamps is 10 min
series3 = pd.date_range('2010-04-15', '2010-05-01', freq='H').union(pd.date_range('2010-05-01 00:10:00', periods=20,
freq='H'))
with warnings.catch_warnings(record=True) as w:
assert bw._get_data_resolution(series3).seconds == 3600
assert len(w) == 1
def test_twelve_by_24():
df = bw.load_campbell_scientific(
bw.datasets.demo_campbell_scientific_site_data)
graph, table12x24 = bw.twelve_by_24(df.Spd40mN,
var_name_label='wind speed',
return_data=True)
graph = bw.twelve_by_24(df.PrcpTot, aggregation_method='sum')
def custom_agg(x):
return x.mean() + (2 * x.std())
graph, table12x24 = bw.twelve_by_24(df.PrcpTot,
aggregation_method=custom_agg,
return_data=True)
assert True
def test_calc_air_density():
data = bw.load_campbell_scientific(bw.datasets.demo_campbell_scientific_site_data)
bw.calc_air_density(data[['T2m']], data[['P2m']])
bw.calc_air_density(data.T2m, data.P2m)
bw.calc_air_density(data.T2m, data.P2m, elevation_ref=0, elevation_site=200)
with pytest.raises(TypeError) as except_info:
bw.calc_air_density(15, 1013, elevation_site=200)
assert str(except_info.value) == 'elevation_ref should be a number'
with pytest.raises(TypeError) as except_info:
bw.calc_air_density(15, 1013, elevation_ref=200)
assert str(except_info.value) == 'elevation_site should be a number'
assert abs(bw.calc_air_density(15, 1013) - 1.225) < 1e-3
assert abs(bw.calc_air_density(15, 1013, elevation_ref=0, elevation_site=200) - 1.203) < 1e-3
assert (abs(bw.calc_air_density(pd.Series([15, 12.5, -5, 23]), pd.Series([1013, 990, 1020, 900])) -
pd.Series([1.225, 1.208, 1.326, 1.059])) < 1e-3).all()
def test_sector_ratio_by_sector():
data = bw.load_csv(bw.datasets.shell_flats_80m_csv)
bw.SectorRatio.by_sector(data['WS70mA100NW_Avg'],
data['WS70mA100SE_Avg'],
data['WD50mW200PNW_VAvg'],
sectors=72,
boom_dir_1=315,
boom_dir_2=135,
return_data=True)[1]
data = bw.load_campbell_scientific(bw.datasets.demo_site_data)
bw.SectorRatio.by_sector(data.Spd40mN,
data.Spd60mN,
wdir=data.Dir38mS,
direction_bin_array=[0, 45, 135, 180, 220, 360],
boom_dir_1=160,
boom_dir_2=340)
assert True
def test_export_tab_file():
df = bw.load_campbell_scientific(
bw.datasets.demo_campbell_scientific_site_data)
graph, tab = bw.freq_table(df.Spd40mN, df.Dir38mS, return_data=True)
bw.export_tab_file(tab, 80, 10, 10, file_name='export_tab_file.tab')
def test_export_tab_file():
df = bw.load_campbell_scientific(
bw.datasets.demo_campbell_scientific_site_data)
graph, tab = bw.freq_table(df.Spd40mN, df.Dir38mS, return_data=True)
bw.export_tab_file(tab, name='campbell_tab_file', lat=10, long=10)
def test_offset_timestamps():
series1 = bw.load_campbell_scientific(
bw.datasets.demo_campbell_scientific_site_data)
# sending index with no start end
bw.offset_timestamps(series1.index, offset='90min')
# sending index with start end
op = bw.offset_timestamps(series1.index,
offset='2min',
date_from='2016-01-01 00:10:00')
assert op[0] == pd.to_datetime('2016-01-01 00:00:00')
assert op[1] == pd.to_datetime('2016-01-01 00:12:00')
op = bw.offset_timestamps(series1.index,
'2min',
date_to='2016-01-01 00:30:00')
assert op[3] == pd.to_datetime('2016-01-01 00:32:00')
assert op[4] == pd.to_datetime('2016-01-01 00:40:00')
op = bw.offset_timestamps(series1.index,
'3min',
date_from='2016-01-01 00:10:00',
date_to='2016-01-01 00:30:00')
assert op[0] == pd.to_datetime('2016-01-01 00:00:00')
assert op[1] == pd.to_datetime('2016-01-01 00:13:00')
assert op[5] == pd.to_datetime('2016-01-01 00:50:00')
op = bw.offset_timestamps(series1.index,
'10min',
date_from='2016-01-01 00:10:00',
date_to='2016-01-01 00:30:00')
assert op[0] == series1.index[0]
assert op[1] == series1.index[2]
# sending DataFrame with datetime index
op = bw.offset_timestamps(series1,
offset='-10min',
date_from='2016-01-01 00:20:00')
assert (op.iloc[1] == series1.iloc[1]).all()
assert len(op) + 1 == len(series1)
assert (op.loc['2016-01-01 00:40:00'] == series1.loc['2016-01-01 00:50:00']
).all()
op = bw.offset_timestamps(series1,
offset='-10min',
date_from='2016-01-01 00:20:00',
overwrite=True)
assert (op.loc['2016-01-01 00:10:00'] == series1.loc['2016-01-01 00:20:00']
).all()
op = bw.offset_timestamps(series1,
'10min',
date_from='2016-01-01 00:10:00',
date_to='2016-01-01 00:30:00')
assert (op.loc['2016-01-01 00:20:00'] == series1.loc['2016-01-01 00:10:00']
).all()
assert (op.loc['2016-01-01 00:40:00'] == series1.loc['2016-01-01 00:40:00']
).all()
assert len(op) + 1 == len(series1)
op = bw.offset_timestamps(series1,
'10min',
date_from='2016-01-01 00:10:00',
date_to='2016-01-01 00:30:00',
overwrite=True)
assert (op.loc['2016-01-01 00:40:00'] == series1.loc['2016-01-01 00:30:00']
).all()
assert len(op) + 1 == len(series1)
# sending Series with datetime index
op = bw.offset_timestamps(series1.Spd60mN,
offset='-10min',
date_from='2016-01-01 00:20:00')
assert (op.iloc[1] == series1.Spd60mN.iloc[1]).all()
assert len(op) + 1 == len(series1.Spd60mN)
assert (op.loc['2016-01-01 00:40:00'] ==
series1.Spd60mN.loc['2016-01-01 00:50:00']).all()
op = bw.offset_timestamps(series1.Spd60mN,
offset='-10min',
date_from='2016-01-01 00:20:00',
overwrite=True)
assert (op.loc['2016-01-01 00:10:00'] ==
series1.Spd60mN.loc['2016-01-01 00:20:00']).all()
op = bw.offset_timestamps(series1.Spd60mN,
'10min',
date_from='2016-01-01 00:10:00',
date_to='2016-01-01 00:30:00')
assert (op.loc['2016-01-01 00:20:00'] ==
series1.Spd60mN.loc['2016-01-01 00:10:00']).all()
assert (op.loc['2016-01-01 00:40:00'] ==
series1.Spd60mN.loc['2016-01-01 00:40:00']).all()
assert len(op) + 1 == len(series1.Spd60mN)
op = bw.offset_timestamps(series1.Spd60mN,
'10min',
date_from='2016-01-01 00:10:00',
date_to='2016-01-01 00:30:00',
overwrite=True)
assert (op.loc['2016-01-01 00:40:00'] ==
series1.Spd60mN.loc['2016-01-01 00:30:00']).all()
assert len(op) + 1 == len(series1.Spd60mN)
