Skip to content

Commit

Permalink
immproved inefficiency in fct group_frequency
Browse files Browse the repository at this point in the history
  • Loading branch information
@ValentinGebhart
ValentinGebhart committed Oct 4, 2024
1 parent e608755 commit 9edac4d
Show file tree
Hide file tree
Showing 2 changed files with 70 additions and 43 deletions.
40 changes: 30 additions & 10 deletions climada/util/interpolation.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -344,16 +344,36 @@ def group_frequency(frequency, value, n_sig_dig=2):
if frequency.size == 0 and value.size == 0:
return ([], [])

if len(value) != len(np.unique(sig_dig_list(value, n_sig_dig=n_sig_dig))):
#check ordering of value
if not all(sorted(value) == value):
# round values and group them
value = round_to_sig_digits(value, n_sig_dig)
value_unique, start_indices = np.unique(value, return_index=True)

if value_unique.size != frequency.size:
if not all(sorted(start_indices) == start_indices):
raise ValueError('Value array must be sorted in ascending order.')
# add frequency for equal value
value, start_indices = np.unique(
sig_dig_list(value, n_sig_dig=n_sig_dig), return_index=True)
start_indices = np.insert(start_indices, len(value), len(frequency))
frequency = np.array([
sum(frequency[start_indices[i]:start_indices[i+1]])
for i in range(len(value))
])
start_indices = np.insert(start_indices, value_unique.size, frequency.size)
frequency = np.add.reduceat(frequency, start_indices[:-1])
return frequency, value_unique

return frequency, value

def round_to_sig_digits(x, n_sig_dig):
"""round each element array to a number of significant digits
Parameters
----------
x : array-like
array to be rounded
n_sig_dig : int
number of significant digits.
Returns
-------
np.array
rounded array
"""
x = np.asarray(x)
x_positive = np.where(np.isfinite(x) & (x != 0), np.abs(x), 10**(n_sig_dig-1))
mags = 10 ** (n_sig_dig - 1 - np.floor(np.log10(x_positive)))
return np.round(x * mags) / mags
73 changes: 40 additions & 33 deletions climada/util/test/test_interpolation.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -22,7 +22,7 @@
import unittest
import numpy as np

from climada.util.interpolation import interpolate_ev, stepfunction_ev, group_frequency, preprocess_and_interpolate_ev
import climada.util.interpolation as u_interp


class TestFitMethods(unittest.TestCase):
Expand All @@ -34,15 +34,15 @@ def test_interpolate_ev_linear_interp(self):
y_train = np.array([8., 4., 2.])
x_test = np.array([0., 3., 4., 6.])
np.testing.assert_allclose(
interpolate_ev(x_test, x_train, y_train),
u_interp.interpolate_ev(x_test, x_train, y_train),
np.array([np.nan, 4., 3., np.nan])
)
np.testing.assert_allclose(
interpolate_ev(x_test, x_train, y_train, extrapolation='extrapolate_constant'),
u_interp.interpolate_ev(x_test, x_train, y_train, extrapolation='extrapolate_constant'),
np.array([8., 4., 3., np.nan])
)
np.testing.assert_allclose(
interpolate_ev(x_test, x_train, y_train,
u_interp.interpolate_ev(x_test, x_train, y_train,
extrapolation='extrapolate_constant', y_asymptotic = 0),
np.array([8., 4., 3., 0.])
)
Expand All @@ -53,16 +53,16 @@ def test_interpolate_ev_threshold_parameters(self):
y_train = np.array([4., 1., 4.])
x_test = np.array([-1., 3., 4.])
np.testing.assert_allclose(
interpolate_ev(x_test, x_train, y_train, extrapolation='extrapolate_constant'),
u_interp.interpolate_ev(x_test, x_train, y_train, extrapolation='extrapolate_constant'),
np.array([4., 1., 2.])
)
np.testing.assert_allclose(
interpolate_ev(x_test, x_train, y_train, x_threshold=1.,
u_interp.interpolate_ev(x_test, x_train, y_train, x_threshold=1.,
extrapolation='extrapolate_constant'),
np.array([1., 1., 2.])
)
np.testing.assert_allclose(
interpolate_ev(x_test, x_train, y_train, y_threshold=2.,
u_interp.interpolate_ev(x_test, x_train, y_train, y_threshold=2.,
extrapolation='extrapolate_constant'),
np.array([4., 4., 4.])
)
Expand All @@ -73,26 +73,26 @@ def test_interpolate_ev_scale_parameters(self):
y_train = np.array([1., 3.])
x_test = np.array([1e0, 1e2])
np.testing.assert_allclose(
interpolate_ev(x_test, x_train, y_train, logx=True, extrapolation='extrapolate'),
u_interp.interpolate_ev(x_test, x_train, y_train, logx=True, extrapolation='extrapolate'),
np.array([0., 2.])
)
np.testing.assert_allclose(
interpolate_ev(x_test, x_train, y_train, logx=True,
u_interp.interpolate_ev(x_test, x_train, y_train, logx=True,
extrapolation='extrapolate_constant'),
np.array([1., 2.])
)
x_train = np.array([1., 3.])
y_train = np.array([1e1, 1e3])
x_test = np.array([0., 2.])
np.testing.assert_allclose(
interpolate_ev(x_test, x_train, y_train, logy=True, extrapolation='extrapolate'),
u_interp.interpolate_ev(x_test, x_train, y_train, logy=True, extrapolation='extrapolate'),
np.array([1e0, 1e2])
)
x_train = np.array([1e1, 1e3])
y_train = np.array([1e1, 1e5])
x_test = np.array([1e0, 1e2])
np.testing.assert_allclose(
interpolate_ev(x_test, x_train, y_train, logx=True, logy=True,
u_interp.interpolate_ev(x_test, x_train, y_train, logx=True, logy=True,
extrapolation='extrapolate'),
np.array([1e-1, 1e3])
)
Expand All @@ -103,7 +103,7 @@ def test_interpolate_ev_degenerate_input(self):
x_test = np.array([0., 2., 4.])
y_train = np.zeros(3)
np.testing.assert_allclose(
interpolate_ev(x_test, x_train, y_train),
u_interp.interpolate_ev(x_test, x_train, y_train),
np.array([np.nan, 0., 0.])
)

Expand All @@ -113,27 +113,27 @@ def test_interpolate_ev_small_input(self):
y_train = np.array([2.])
x_test = np.array([0., 1., 2.])
np.testing.assert_allclose(
interpolate_ev(x_test, x_train, y_train, extrapolation='extrapolate'),
u_interp.interpolate_ev(x_test, x_train, y_train, extrapolation='extrapolate'),
np.array([2., 2., np.nan])
)
np.testing.assert_allclose(
interpolate_ev(x_test, x_train, y_train, extrapolation='extrapolate', y_asymptotic=0),
u_interp.interpolate_ev(x_test, x_train, y_train, extrapolation='extrapolate', y_asymptotic=0),
np.array([2., 2., 0.])
)
np.testing.assert_allclose(
interpolate_ev(x_test, x_train, y_train),
u_interp.interpolate_ev(x_test, x_train, y_train),
np.full(3, np.nan)
)

x_train = np.array([])
y_train = np.array([])
x_test = np.array([0., 1., 2.])
np.testing.assert_allclose(
interpolate_ev(x_test, x_train, y_train),
u_interp.interpolate_ev(x_test, x_train, y_train),
np.full(3, np.nan)
)
np.testing.assert_allclose(
interpolate_ev(x_test, x_train, y_train,
u_interp.interpolate_ev(x_test, x_train, y_train,
extrapolation='extrapolate_constant', y_asymptotic=0),
np.zeros(3)
)
Expand All @@ -144,11 +144,11 @@ def test_stepfunction_ev(self):
y_train = np.array([8., 4., 2.])
x_test = np.array([0., 3., 4., 6.])
np.testing.assert_allclose(
stepfunction_ev(x_test, x_train, y_train),
u_interp.stepfunction_ev(x_test, x_train, y_train),
np.array([8., 4., 2., np.nan])
)
np.testing.assert_allclose(
stepfunction_ev(x_test, x_train, y_train, y_asymptotic=0.),
u_interp.stepfunction_ev(x_test, x_train, y_train, y_asymptotic=0.),
np.array([8., 4., 2., 0.])
)

Expand All @@ -158,39 +158,46 @@ def test_stepfunction_ev_small_input(self):
y_train = np.array([2.])
x_test = np.array([0., 1., 2.])
np.testing.assert_allclose(
stepfunction_ev(x_test, x_train, y_train),
u_interp.stepfunction_ev(x_test, x_train, y_train),
np.array([2., 2., np.nan])
)
np.testing.assert_allclose(
stepfunction_ev(x_test, x_train, y_train, y_asymptotic=0),
u_interp.stepfunction_ev(x_test, x_train, y_train, y_asymptotic=0),
np.array([2., 2., 0.])
)
x_train = np.array([])
y_train = np.array([])
x_test = np.array([0., 1., 2.])
np.testing.assert_allclose(
stepfunction_ev(x_test, x_train, y_train),
u_interp.stepfunction_ev(x_test, x_train, y_train),
np.full(3, np.nan)
)
np.testing.assert_allclose(
stepfunction_ev(x_test, x_train, y_train, y_asymptotic=0),
u_interp.stepfunction_ev(x_test, x_train, y_train, y_asymptotic=0),
np.zeros(3)
)

def test_frequency_group(self):
"""Test frequency grouping method"""
frequency = np.ones(6)
intensity = np.array([1., 1., 1., 2., 3., 3])
intensity = np.array([1.00001, .999, 1., 2., 3., 3])
np.testing.assert_allclose(
group_frequency(frequency, intensity),
u_interp.group_frequency(frequency, intensity),
([3, 1, 2], [1, 2, 3])
)
np.testing.assert_allclose(
group_frequency([], []),
u_interp.group_frequency([], []),
([], [])
)
with self.assertRaises(ValueError):
group_frequency(frequency, intensity[::-1])
u_interp.group_frequency(frequency, intensity[::-1])

def test_round_to_sig_digits(self):
array = [.00111, 999., 55.5, 0., -1.001, -1.08]
np.testing.assert_allclose(
u_interp.round_to_sig_digits(array, n_sig_dig=2),
[.0011, 1000., 56, 0., -1., -1.1]
)

def test_preprocess_and_interpolate_ev(self):
"""Test wrapper function"""
Expand All @@ -202,30 +209,30 @@ def test_preprocess_and_interpolate_ev(self):
# test interpolation
np.testing.assert_allclose(
[np.nan, 55., np.nan],
preprocess_and_interpolate_ev(test_frequency, None, frequency, values)
u_interp.preprocess_and_interpolate_ev(test_frequency, None, frequency, values)
)
np.testing.assert_allclose(
[np.nan, .55, np.nan],
preprocess_and_interpolate_ev(None, test_values, frequency, values)
u_interp.preprocess_and_interpolate_ev(None, test_values, frequency, values)
)

# test extrapolation with constants
np.testing.assert_allclose(
[100. , 55., 0.],
preprocess_and_interpolate_ev(test_frequency, None, frequency, values,
u_interp.preprocess_and_interpolate_ev(test_frequency, None, frequency, values,
method='extrapolate_constant', y_asymptotic=0.)
)
np.testing.assert_allclose(
[1., .55, np.nan],
preprocess_and_interpolate_ev(None, test_values, frequency, values,
u_interp.preprocess_and_interpolate_ev(None, test_values, frequency, values,
method='extrapolate_constant')
)

# test error raising
with self.assertRaises(ValueError):
preprocess_and_interpolate_ev(test_frequency, test_values, frequency, values)
u_interp.preprocess_and_interpolate_ev(test_frequency, test_values, frequency, values)
with self.assertRaises(ValueError):
preprocess_and_interpolate_ev(None, None, frequency, values)
u_interp.preprocess_and_interpolate_ev(None, None, frequency, values)

# Execute Tests
if __name__ == "__main__":
Expand Down

0 comments on commit 9edac4d

Please sign in to comment.