def test_lnqmed_residuals_two_donors(self):
analysis = QmedAnalysis(self.catchment, CatchmentCollections(self.db_session), year=2000)
donors = analysis.find_donor_catchments()[0:2] # 17001, 10001
qmed_amax = [QmedAnalysis(d).qmed() for d in donors]
qmed_descr =[QmedAnalysis(d, year=2000).qmed(method='descriptors') for d in donors]
assert_almost_equal(qmed_amax, [90.532, 50.18]) # not verified
assert_almost_equal(qmed_descr, [51.73180402, 48.70106637]) # not verified
result = [analysis._lnqmed_residual(d) for d in donors]
assert_almost_equal(result, [0.55963062, 0.02991561])
def test_vector_b_two_donors(self):
analysis = QmedAnalysis(self.catchment, CatchmentCollections(self.db_session), year=2000)
donors = analysis.find_donor_catchments()[0:2] # 17001, 10001
# [0.352256808, 0.002198921] * 0.1175 = [0.041390175, 0.000258373]
result = analysis._vec_b(donors)
assert_almost_equal(result, [0.041390175, 0.000258373])
def test_pot_2_years(self):
catchment = Catchment("Aberdeen", "River Dee")
catchment.pot_dataset = PotDataset(start_date=date(1998, 1, 1), end_date=date(1999, 12, 31))
catchment.pot_dataset.pot_records = [PotRecord(date(1999, 1, 1), 3.0, 0.5),
PotRecord(date(1999, 2, 1), 2.0, 0.5),
PotRecord(date(1999, 12, 31), 1.0, 0.5)]
self.assertAlmostEqual(QmedAnalysis(catchment).qmed(method='pot_records'), 1.8789, 4)
def test_matrix_sigma_eta_two_donors(self):
analysis = QmedAnalysis(self.catchment, CatchmentCollections(self.db_session), year=2000)
donors = analysis.find_donor_catchments()[0:2] # 17001, 10001
result = analysis._matrix_sigma_eta(donors)
# 0.1175 * 0.001908936 = 0.00022430
assert_almost_equal(result, [[0.1175, 0.00022430],
[0.00022430, 0.1175]])
def test_amax_long_records(self):
catchment = Catchment("Aberdeen", "River Dee")
catchment.amax_records = [AmaxRecord(date(1999, 12, 31), 5.0, 0.5),
AmaxRecord(date(2000, 12, 31), 1.0, 0.5),
AmaxRecord(date(2001, 12, 31), 4.0, 0.5),
AmaxRecord(date(2002, 12, 31), 2.0, 0.5),
AmaxRecord(date(2003, 12, 31), 3.0, 0.5)]
self.assertEqual(QmedAnalysis(catchment).qmed(method='amax_records'), 3)
def test_descriptors_2008_urban_adjustment(self):
catchment = Catchment("Aberdeen", "River Dee")
catchment.descriptors = Descriptors(dtm_area=1,
bfihost=0.50,
sprhost=50,
saar=1000,
farl=1,
urbext2000=1)
self.assertAlmostEqual(QmedAnalysis(catchment, year=2000).urban_adj_factor(), 2.970205798, 4)
def test_qmed_two_donors(self):
analysis = QmedAnalysis(self.catchment, CatchmentCollections(self.db_session), year=2000)
donors = analysis.find_donor_catchments()[0:2] # 17001, 10001
result = analysis.qmed(method='descriptors', donor_catchments=donors)
# exp(ln(0.61732109) + 0.34379622 * 0.55963062 + 0.00102012 * 0.02991561) =
# exp(ln(0.61732109) + 0.192429411) = 0.748311028
self.assertAlmostEqual(result, 0.748311028, places=5)
def test_descriptors_1999_2(self):
catchment = Catchment("Aberdeen", "River Dee")
catchment.descriptors = Descriptors(dtm_area=2.345,
bfihost=0,
sprhost=100,
saar=2000,
farl=0.5,
urbext2000=0)
self.assertAlmostEqual(QmedAnalysis(catchment).qmed(method='descriptors_1999'), 0.3729, 4)
def test_descriptors_2008_1(self):
catchment = Catchment("Aberdeen", "River Dee")
catchment.descriptors = Descriptors(dtm_area=1,
bfihost=0.50,
sprhost=50,
saar=1000,
farl=1,
urbext2000=0)
self.assertAlmostEqual(QmedAnalysis(catchment).qmed(method='descriptors_2008'), 0.5907, 4)
def test_model_error_corr_between_two_donors(self):
analysis = QmedAnalysis(self.catchment, CatchmentCollections(self.db_session), year=2000)
donors = analysis.find_donor_catchments()[0:2] # 17001, 10001
dist = donors[0].distance_to(donors[1])
self.assertAlmostEqual(dist, 188.8487072) # not verified
result = analysis._model_error_corr(donors[0], donors[1])
self.assertAlmostEqual(result, 0.001908936)
def test_descriptors_2008_urban(self):
catchment = Catchment("Aberdeen", "River Dee")
catchment.descriptors = Descriptors(dtm_area=1,
bfihost=0.50,
sprhost=50,
saar=1000,
farl=1,
urbext2000=1)
self.assertAlmostEqual(QmedAnalysis(catchment, year=2000).qmed(method='descriptors_2008', as_rural=False),
1.7546, 4)
def _run_qmed_analysis(self):
results = {}
analysis = QmedAnalysis(self.catchment,
self.gauged_catchments,
results_log=results)
self.qmed = analysis.qmed()
results['qmed'] = self.qmed
self.results['qmed'] = results
def test_pot_records_by_month(self):
catchment = Catchment("Aberdeen", "River Dee")
catchment.pot_dataset = PotDataset(start_date=date(1998, 10, 1), end_date=date(2000, 1, 31))
catchment.pot_dataset.pot_records = [PotRecord(date(1999, 1, 1), 3.0, 0.5),
PotRecord(date(1999, 2, 1), 2.0, 0.5),
PotRecord(date(1999, 2, 15), 2.0, 0.5),
PotRecord(date(1999, 12, 31), 1.0, 0.5)]
analysis = QmedAnalysis(catchment)
records_by_month = analysis._pot_month_counts(catchment.pot_dataset)
expected = [2, 1, 1, 1, 1, 1, 1, 1, 1, 2, 2, 2]
result = [len(month) for month in records_by_month]
self.assertEqual(result, expected)
def test_pot_complete_years(self):
catchment = Catchment("Aberdeen", "River Dee")
catchment.pot_dataset = PotDataset(start_date=date(1998, 10, 1), end_date=date(2000, 1, 31))
catchment.pot_dataset.pot_records = [PotRecord(date(1999, 1, 1), 3.0, 0.5),
PotRecord(date(1999, 2, 1), 2.0, 0.5),
PotRecord(date(1999, 2, 15), 2.0, 0.5),
PotRecord(date(1999, 12, 31), 1.0, 0.5)]
analysis = QmedAnalysis(catchment)
records, n = analysis._complete_pot_years(catchment.pot_dataset)
result = [record.date for record in records]
expected = [date(1999, 2, 1),
date(1999, 2, 15),
date(1999, 12, 31)]
self.assertEqual(result, expected)
def test_matrix_sigma_eps_two_donors(self):
analysis = QmedAnalysis(self.catchment, CatchmentCollections(self.db_session), year=2000)
donors = analysis.find_donor_catchments()[0:2] # 17001, 10001
record0 = [donors[0].amax_records_start(), donors[0].amax_records_end()]
record1 = [donors[1].amax_records_start(), donors[1].amax_records_end()]
self.assertEqual(record0, [1969, 2005]) # n=37
self.assertEqual(record1, [1939, 1984]) # n=46, 16 years overlapping
result = analysis._matrix_sigma_eps(donors)
# 4 * 0.16351290**2 / 37 = 0.00289043
# 4 * 0.16351290 * 0.20423656 * 16 / 37 / 46 * 0.133632774 = 0.00016781
# 4 * 0.20423656**2 / 46 = 0.00362718
assert_almost_equal(result, [[0.00289043, 0.00016781],
[0.00016781, 0.00362718]])
def test_all(self):
catchment = Catchment("Aberdeen", "River Dee")
catchment.channel_width = 1
catchment.amax_records = [AmaxRecord(date(1999, 12, 31), 1.0, 0.5),
AmaxRecord(date(2000, 12, 31), 1.0, 0.5)]
catchment.descriptors = Descriptors(dtm_area=1,
bfihost=0.50,
sprhost=50,
saar=1000,
farl=1,
urbext2000=0)
qmeds = QmedAnalysis(catchment).qmed_all_methods()
self.assertEqual(qmeds['amax_records'], 1)
self.assertEqual(qmeds['channel_width'], 0.182)
self.assertEqual(qmeds['area'], 1.172)
self.assertAlmostEqual(qmeds['descriptors_1999'], 0.2671, 4)
def analyse_catchment(catchment, gauged_catchments):
result = {}
result['id'] = catchment.id
analysis = QmedAnalysis(catchment)
result['qmed_amax'] = analysis.qmed(method='amax_records')
result['qmed_descr'] = analysis.qmed(method='descriptors')
result['qmed_descr_1999'] = analysis.qmed(method='descriptors_1999')
analysis.gauged_catchments = gauged_catchments
donors = analysis.find_donor_catchments()
analysis.idw_power = 2
result['qmed_descr_idw'] = analysis.qmed(method='descriptors',
donor_catchments=donors)
analysis.idw_power = 3
result['qmed_descr_idw3'] = analysis.qmed(method='descriptors',
donor_catchments=donors)
analysis.donor_weighting = 'equal'
result['qmed_descr_first'] = analysis.qmed(method='descriptors',
donor_catchments=donors[0:1])
return result
def run_qmed_analysis(self):
self.qmed_analysis = QmedAnalysis(self.catchment,
self.gauged_catchments)
self.results = self.qmed_analysis.results_log
self.results['qmed_all_methods'] = self.qmed_analysis.qmed_all_methods(
)
def test_no_descriptors_1999(self):
catchment = Catchment("Aberdeen", "River Dee")
try:
QmedAnalysis(catchment).qmed(method='descriptors_1999')
except InsufficientDataError as e:
self.assertEqual(str(e), "Catchment `descriptors` attribute must be set first.")
def test_vector_alpha_two_donors(self):
analysis = QmedAnalysis(self.catchment, CatchmentCollections(self.db_session), year=2000)
donors = analysis.find_donor_catchments()[0:2] # 17001, 10001
result = analysis._vec_alpha(donors)
assert_almost_equal(result, [0.34379622, 0.00102012]) # calculated in Excel
def test_distance_two_donors(self):
analysis = QmedAnalysis(self.catchment, CatchmentCollections(self.db_session), year=2000)
donors = analysis.find_donor_catchments()[0:2] # 17001, 10001
result = [d.distance_to(self.catchment) for d in donors]
assert_almost_equal(result, [5, 183.8515], decimal=4)
def test_lnqmed_corr(self):
analysis = QmedAnalysis(self.catchment, CatchmentCollections(self.db_session), year=2000)
donors = analysis.find_donor_catchments()[0:2] # 17001, 10001
results = analysis._lnqmed_corr(donors[0], donors[1])
self.assertAlmostEqual(results, 0.133632774)
def test_beta_two_donors(self):
analysis = QmedAnalysis(self.catchment, CatchmentCollections(self.db_session), year=2000)
donors = analysis.find_donor_catchments()[0:2] # 17001, 10001
result = [analysis._beta(d) for d in donors]
assert_almost_equal(result, [0.16351290, 0.20423656])
def test_matrix_omega_one_donor(self):
# 0.1175 + 0.18292242 = 0.30042242
result = QmedAnalysis(self.catchment)._matrix_omega([self.donor_catchment])
assert_almost_equal(result, [[0.30042242]])
def test_vector_alpha_one_donor(self):
# 1/0.30042242 * 0.1175 = 0.39111595
result = QmedAnalysis(self.catchment)._vec_alpha([self.donor_catchment])
assert_almost_equal(result, [0.39111595])
def test_qmed_one_donor(self):
# 0.61732109 * 1.69282714**0.39111595 = 0.75844685
result = QmedAnalysis(self.catchment).qmed(method='descriptors', donor_catchments=[self.donor_catchment])
self.assertAlmostEqual(result, 0.75844685, places=4)
def test_model_error_corr_two_donors(self):
analysis = QmedAnalysis(self.catchment, CatchmentCollections(self.db_session), year=2000)
donors = analysis.find_donor_catchments()[0:2] # 17001, 10001
result = [analysis._model_error_corr(self.catchment, d) for d in donors]
assert_almost_equal(result, [0.352256808, 0.002198921])
from floodestimation.loaders import load_catchment
from floodestimation import db
from floodestimation.collections import CatchmentCollections
from floodestimation.analysis import QmedAnalysis
db_session = db.Session()
dee_catchment = load_catchment('nith_cds.cd3')
gauged_catchments = CatchmentCollections(db_session)
qmed_analysis = QmedAnalysis(dee_catchment, gauged_catchments)
print(qmed_analysis.qmed())
print(qmed_analysis.methods)
print(qmed_analysis.qmed_all_methods())
print(qmed_analysis.urban_adj_factor())
print(qmed_analysis.find_donor_catchments(5, 200.0))
qmed_analysis.idw_power = 1.5
print(qmed_analysis.idw_power)
donors = qmed_analysis.find_donor_catchments(5, 200.0)
for donor in donors:
Q = QmedAnalysis(donors[0], gauged_catchments)
print(donor, qmed_analysis._error_correlation(donor), Q.qmed_all_methods())
db_session.close()
def test_matrix_omega_two_donors(self):
analysis = QmedAnalysis(self.catchment, CatchmentCollections(self.db_session), year=2000)
donors = analysis.find_donor_catchments()[0:2] # 17001, 10001
result = analysis._matrix_omega(donors)
assert_almost_equal(result, [[0.12039043, 0.00039211],
[0.00039211, 0.12112718]])
def test_area_3(self):
catchment = Catchment("Aberdeen", "River Dee")
catchment.descriptors.dtm_area = 100
self.assertAlmostEqual(QmedAnalysis(catchment).qmed(method='area'), 81.2790, 4)
