def recalculateRefPSI(col2psi_str, col2total_count, col2weights):
"""
Reference is assumed to be col 0
"""
psi_vals = []
total_vals = []
weights = []
for col in col2psi_str:
if col == 0:
continue
if col2psi_str[col] == NA:
continue
psi_vals.append(float(col2psi_str[col]))
total_vals.append(col2total_count[col])
if col2weights:
weights.append(col2weights[col - 1])
if col2weights:
# TODO Pretty sure int(round(x)) == round(x)
median_psi = robustats.weighted_median(psi_vals, weights)
median_total = int(
round(robustats.weighted_median(total_vals, weights)))
else:
median_psi = statistics.median(psi_vals)
try:
median_total = int(round(statistics.median(total_vals)))
except:
pdb.set_trace()
return "%.2f" % median_psi, median_total
def test_generic_4(self):
x = [
0.49, 0.36, 0.36, 0.18, 0.75, 0.33, 0.68, 0.82, 0.38, 0.75, 0.61,
0.02, 0.57, 0.23, 0.56, 0.03, 0.45, 0.44, 0.36, 0.92
]
weights = [
0.08, 0.22, 0.79, 0.84, 0.69, 0.84, 0.08, 0.87, 0.95, 0.27, 0.9,
0.34, 0.75, 0.65, 0.02, 0.83, 0.32, 0.68, 0.92, 0.37
]
weighted_median = robustats.weighted_median(x, weights)
self.assertEqual(weighted_median, 0.38)
def test_generic_5(self):
x = [
0.64, 0.95, 0.05, 0.08, 0.32, 0.25, 0.58, 0.69, 0.88, 0.53, 0.48,
0.58, 0.32, 0.52, 0.42, 0.69, 0.43, 0.91, 0.15, 0.27, 0.31, 0.16,
0.56, 0.68, 0.58, 0.04, 0.51, 0.06, 0.18, 0.03
]
weights = [
0.97, 0.2, 0.12, 0.01, 0.86, 0.29, 0.93, 0.96, 0.89, 0.03, 0.24,
0.56, 0.81, 0.97, 0.48, 0.32, 0.33, 0.22, 0.8, 0.17, 0.96, 0.75,
0.43, 0.24, 0.81, 0.4, 0.93, 0.43, 0.17, 0.13
]
weighted_median = robustats.weighted_median(x, weights)
self.assertEqual(weighted_median, 0.51)
def test_generic_6(self):
x = [
0.19, 0.14, 0.15, 0.79, 0.36, 0.13, 0.44, 0.67, 0.44, 0.98, 0.2,
0.11, 0.78, 0.67, 0.28, 0.29, 0.99, 0.55, 0.34, 0.36, 0.09, 0.13,
0.56, 0.19, 0.08, 0.46, 0.62, 0.98, 0.46, 0.37, 0.09, 0.94, 0.84,
0.64, 0.18, 0.64, 0.78, 0.88, 0.17, 0.28
]
weights = [
0.67, 0.39, 0.31, 0.06, 0.93, 0.21, 0.09, 0.29, 0.78, 0.42, 0.79,
0.27, 0.77, 0.35, 0.11, 0.99, 0.05, 0.39, 0.34, 0.97, 0.82, 0.4,
0.09, 0.77, 0.28, 0.03, 0.63, 0.67, 0.1, 0.3, 0.85, 0.44, 0.66,
0.52, 0.15, 0.4, 0.82, 0.66, 0.21, 0.72
]
weighted_median = robustats.weighted_median(x, weights)
self.assertEqual(weighted_median, 0.36)
def recalculateRefPSI_list(psi_list, col2weights):
"""
Reference is assumed to be col 0
"""
vals = []
weights = []
for i in range(1, len(psi_list)):
if psi_list[i] == NA:
continue
vals.append(float(psi_list[i]))
if col2weights:
weights.append(col2weights[i - 1])
if col2weights:
# TODO delete
median_psi = robustats.weighted_median(vals, weights)
else:
median_psi = statistics.median(vals)
return "%.2f" % median_psi
def test_same_weights(self):
x = [1., 2., 3.]
weights = [1., 1., 1.]
weighted_median = robustats.weighted_median(x, weights)
self.assertEqual(weighted_median, 2.)
def test_generic_3(self):
x = [0.1, 0.35, 0.05, 0.1, 0.15, 0.05, 0.2]
weights = [0.1, 0.35, 0.05, 0.1, 0.15, 0.05, 0.2]
weighted_median = robustats.weighted_median(x, weights)
self.assertEqual(weighted_median, 0.2)
def test_generic_2(self):
x = [4.2, 1.3, 7.4, 0.2, 4.6, 9.8, 5.5, 3.7]
weights = [0.4, 2.1, 1.1, 1.6, 0.3, 0.9, 1.2, 1.7]
weighted_median = robustats.weighted_median(x, weights)
self.assertEqual(weighted_median, 3.7)
def test_generic_1(self):
x = [1.3, 5.1, 2.9, 1.9, 7.4]
weights = [1.4, 0.9, 0.6, 1.2, 1.7]
weighted_median = robustats.weighted_median(x, weights)
self.assertEqual(weighted_median, 2.9)
def test_even_list(self):
x = [1., 2.]
weights = [1., 1.]
weighted_median = robustats.weighted_median(x, weights)
self.assertEqual(weighted_median, 1.)
def test_dominant_weight(self):
x = [1., 2., 3.]
weights = [3., 1., 1.]
weighted_median = robustats.weighted_median(x, weights)
self.assertEqual(weighted_median, 1.)
def test_edge_case(self):
x = [1., 2., 3.]
weights = [2., 1., 1.]
weighted_median = robustats.weighted_median(x, weights)
self.assertEqual(weighted_median, 2.)
