def test_local_regressor_cm_intersite(monkeypatch):
"""Test local regressor process for intersite processing"""
from resistics.regression.local import LocalRegressor
from mocks import mock_window_selector
import numpy as np
# patch out the writeTF function
monkeypatch.setattr(LocalRegressor, "writeResult", mock_local_regressor_writeResult)
selector = mock_window_selector()
regressor = LocalRegressor(selector, "test")
regressor.setInput("local", ["Hx", "Hy"], inCross=["Hx", "Hy"])
regressor.setOutput("inter", ["Ex", "Ey"], outCross=["Ex", "Ey"])
regressor.setMethod("cm", intercept=False)
regressor.setSmooth("hann", 1)
regressor.process()
# expected
expected_evalfreq = np.array([24, 40])
expected_impedances = [
np.array([[6.0 + 0.0j, 8.0 + 0.0j], [4.0 + 0.0j, 1.0 + 0.0j]]),
np.array([[6.0 + 0.0j, 8.0 + 0.0j], [4.0 + 0.0j, 1.0 + 0.0j]]),
]
expected_variances = [
np.array([[9.14649740e-29, 5.89343794e-29], [1.42914022e-30, 9.20849678e-31]]),
np.array([[1.38145310e-28, 1.84606779e-28], [3.45363276e-29, 4.61516947e-29]]),
]
np.testing.assert_equal(regressor.evalFreq, expected_evalfreq)
np.testing.assert_almost_equal(regressor.parameters[0], expected_impedances[0])
np.testing.assert_almost_equal(regressor.parameters[1], expected_impedances[1])
np.testing.assert_almost_equal(regressor.variances[0], expected_variances[0])
np.testing.assert_almost_equal(regressor.variances[1], expected_variances[1])
def test_local_regressor_cm(monkeypatch):
"""Test local regressor process using standard parameters"""
from resistics.regression.local import LocalRegressor
from mocks import mock_window_selector
import numpy as np
# patch out the writeTF function
monkeypatch.setattr(LocalRegressor, "writeResult", mock_local_regressor_writeResult)
selector = mock_window_selector()
regressor = LocalRegressor(selector, "test")
regressor.setInput("local", ["Hx", "Hy"])
regressor.setOutput("local", ["Ex", "Ey"])
regressor.setMethod("cm", intercept=False)
regressor.setSmooth("hann", 1)
assert regressor.inCross == ["Hx", "Hy"]
assert regressor.outCross == ["Ex", "Ey"]
regressor.process()
# expected
expected_evalfreq = np.array([24, 40])
expected_impedances = [
np.array([[3 + 0.0j, 5 + 0.0j], [2 + 0.0j, 7 + 0.0j]]),
np.array([[3 + 0.0j, 5 + 0.0j], [2 + 0.0j, 7 + 0.0j]]),
]
expected_variances = [
np.array([[2.08722609e-28, 1.34197332e-28], [2.31914010e-29, 1.49108146e-29]]),
np.array([[4.59299270e-29, 6.12727229e-29], [1.34977745e-28, 1.80066778e-28]]),
]
np.testing.assert_equal(regressor.evalFreq, expected_evalfreq)
np.testing.assert_almost_equal(regressor.parameters[0], expected_impedances[0])
np.testing.assert_almost_equal(regressor.parameters[1], expected_impedances[1])
np.testing.assert_almost_equal(regressor.variances[0], expected_variances[0])
np.testing.assert_almost_equal(regressor.variances[1], expected_variances[1])
def test_local_regressor_cm_tipper(monkeypatch):
"""Test local regressor process for different combination of inputs and ouputs (tipper)"""
from resistics.regression.local import LocalRegressor
from mocks import mock_window_selector
import numpy as np
# patch out the writeTF function
monkeypatch.setattr(LocalRegressor, "writeResult", mock_local_regressor_writeResult)
selector = mock_window_selector()
regressor = LocalRegressor(selector, "test")
regressor.setInput("local", ["Hx", "Hy"], inCross=["Hx", "Hy"])
regressor.setOutput("local", ["Hz"], outCross=None)
regressor.setMethod("cm", intercept=False)
regressor.setSmooth("hann", 1)
regressor.process()
# expected
expected_evalfreq = np.array([24, 40])
expected_impedances = [
np.array([[2.0 + 0.0j, 8.0 + 0.0j]]),
np.array([[2.0 + 0.0j, 8.0 + 0.0j]]),
]
expected_variances = [
np.array([[3.41232327e-27, 2.19175834e-27]]),
np.array([[5.18941937e-29, 6.91709078e-29]]),
]
np.testing.assert_equal(regressor.evalFreq, expected_evalfreq)
np.testing.assert_almost_equal(regressor.parameters[0], expected_impedances[0])
np.testing.assert_almost_equal(regressor.parameters[1], expected_impedances[1])
np.testing.assert_almost_equal(regressor.variances[0], expected_variances[0])
np.testing.assert_almost_equal(regressor.variances[1], expected_variances[1])
def test_local_regressor_mm_intercept(monkeypatch):
"""Test local regressor process with addition of intercept term"""
from resistics.regression.local import LocalRegressor
from mocks import mock_window_selector
import numpy as np
# patch out the writeTF function
monkeypatch.setattr(LocalRegressor, "writeResult", mock_local_regressor_writeResult)
selector = mock_window_selector(intercept=True)
regressor = LocalRegressor(selector, "test")
regressor.setInput("local", ["Hx", "Hy"], inCross=["Hx", "Hy"])
regressor.setOutput("local", ["Ex", "Ey"], outCross=["Ex", "Ey"])
regressor.setMethod("mm", intercept=True)
regressor.setSmooth("hann", 1)
regressor.process()
# expected
expected_evalfreq = np.array([24, 40])
expected_impedances = [
np.array(
[
[3.0 + 0.0j, 5.0 + 0.0j, 5.0 + 0.0j],
[2.0 + 0.0j, 7.0 + 0.0j, -9.0 + 0.0j],
]
),
np.array(
[
[3.0 + 0.0j, 5.0 + 0.0j, 5.0 + 0.0j],
[2.0 + 0.0j, 7.0 + 0.0j, -9.0 + 0.0j],
]
),
]
expected_variances = [
np.array(
[
[1.21088688e-23, 7.25740911e-24, 3.34802360e-23],
[3.95687739e-23, 2.37154094e-23, 1.09405091e-22],
]
),
np.array(
[
[2.48945809e-27, 3.32569517e-27, 2.95106525e-26],
[1.00812766e-27, 1.34676912e-27, 1.19505948e-26],
]
),
]
print(regressor.parameters)
print(regressor.variances)
np.testing.assert_equal(regressor.evalFreq, expected_evalfreq)
np.testing.assert_almost_equal(regressor.parameters[0], expected_impedances[0])
np.testing.assert_almost_equal(regressor.parameters[1], expected_impedances[1])
np.testing.assert_almost_equal(regressor.variances[0], expected_variances[0])
np.testing.assert_almost_equal(regressor.variances[1], expected_variances[1])
def test_local_regressor_getSmoothLen(monkeypatch):
"""Test local regressor process"""
from resistics.regression.local import LocalRegressor
from mocks import mock_window_selector
import numpy as np
selector = mock_window_selector()
from mocks import mock_window_selector
regressor = LocalRegressor(selector, "test")
regressor.setInput("local", ["Hx", "Hy"])
regressor.setOutput("local", ["Ex", "Ey"])
regressor.setSmooth("hann", 1)
assert regressor.smoothLen == 1
regressor.setSmooth("hann", 12)
assert regressor.getSmoothLen(65) == 13
def test_local_regressor_setSmooth():
"""Test local regressor setSmooth"""
from resistics.regression.local import LocalRegressor
from mocks import mock_window_selector
selector = mock_window_selector()
regressor = LocalRegressor(selector, "test")
assert regressor.smoothFunc == "hann"
assert regressor.smoothLen is None
regressor.setSmooth("hann", 7)
assert regressor.smoothFunc == "hann"
assert regressor.smoothLen == 7
regressor.setSmooth("parzen", 8)
assert regressor.smoothFunc == "parzen"
assert regressor.smoothLen == 8
assert regressor.getSmoothLen(65) == 9
def test_local_regressor_cm_noise(monkeypatch):
"""Test local regressor process with addition of noise"""
from resistics.regression.local import LocalRegressor
from mocks import mock_window_selector
import numpy as np
# patch out the writeTF function
monkeypatch.setattr(LocalRegressor, "writeResult", mock_local_regressor_writeResult)
selector = mock_window_selector(localnoise=True)
regressor = LocalRegressor(selector, "test")
regressor.setInput("local", ["Hx", "Hy"], inCross=["Hx", "Hy"])
regressor.setOutput("local", ["Ex", "Ey"], outCross=["Ex", "Ey"])
regressor.setMethod("cm", intercept=False)
regressor.setSmooth("hann", 1)
regressor.process()
# expected
expected_evalfreq = np.array([24, 40])
expected_impedances = [
np.array(
[
[4.0781224 + 0.0j, 3.567369 + 0.0j],
[1.85376383 + 0.0j, 6.98454182 + 0.0j],
]
),
np.array(
[
[3.79687613 + 0.0j, 3.83744749 + 0.0j],
[4.64021435 + 0.0j, 3.41744201 + 0.0j],
]
),
]
expected_variances = [
np.array([[0.04024341, 0.02563386], [0.0075834, 0.0048304]]),
np.array([[0.00760705, 0.01007484], [2.48409977, 3.28996458]]),
]
np.testing.assert_equal(regressor.evalFreq, expected_evalfreq)
np.testing.assert_almost_equal(regressor.parameters[0], expected_impedances[0])
np.testing.assert_almost_equal(regressor.parameters[1], expected_impedances[1])
np.testing.assert_almost_equal(regressor.variances[0], expected_variances[0])
np.testing.assert_almost_equal(regressor.variances[1], expected_variances[1])
def test_local_regressor_mm_noise(monkeypatch):
"""Test local regressor process with addition of noise"""
from resistics.regression.local import LocalRegressor
from mocks import mock_window_selector
import numpy as np
# patch out the writeTF function
monkeypatch.setattr(LocalRegressor, "writeResult", mock_local_regressor_writeResult)
selector = mock_window_selector(localnoise=True)
regressor = LocalRegressor(selector, "test")
regressor.setInput("local", ["Hx", "Hy"], inCross=["Hx", "Hy"])
regressor.setOutput("local", ["Ex", "Ey"], outCross=["Ex", "Ey"])
regressor.setMethod("mm", intercept=False)
regressor.setSmooth("hann", 1)
regressor.process()
# expected
expected_evalfreq = np.array([24, 40])
expected_impedances = [
np.array(
[
[4.30696333 + 0.0j, 3.38002297 + 0.0j],
[2.5217861 + 0.0j, 6.44672487 + 0.0j],
]
),
np.array(
[
[4.23654503 + 0.0j, 3.33331255 + 0.0j],
[-8.19366818 + 0.0j, 18.06316472 + 0.0j],
]
),
]
expected_variances = [
np.array([[1.72623564e-04, 1.09956096e-04], [1.55939006e-01, 9.93285271e-02]]),
np.array([[0.00691985, 0.00916471], [0.03841539, 0.0508777]]),
]
np.testing.assert_equal(regressor.evalFreq, expected_evalfreq)
np.testing.assert_almost_equal(regressor.parameters[0], expected_impedances[0])
np.testing.assert_almost_equal(regressor.parameters[1], expected_impedances[1])
np.testing.assert_almost_equal(regressor.variances[0], expected_variances[0])
np.testing.assert_almost_equal(regressor.variances[1], expected_variances[1])
def test_local_regressor_ols_noise(monkeypatch):
"""Test local regressor process with addition of noise"""
from resistics.regression.local import LocalRegressor
from mocks import mock_window_selector
import numpy as np
# patch out the writeTF function
monkeypatch.setattr(LocalRegressor, "writeResult", mock_local_regressor_writeResult)
selector = mock_window_selector(localnoise=True)
regressor = LocalRegressor(selector, "test")
regressor.setInput("local", ["Hx", "Hy"], inCross=["Hx", "Hy"])
regressor.setOutput("local", ["Ex", "Ey"], outCross=["Ex", "Ey"])
regressor.setMethod("ols", intercept=False)
regressor.setSmooth("hann", 1)
regressor.process()
# expected
expected_evalfreq = np.array([24, 40])
expected_impedances = [
np.array(
[
[4.28572653 + 0.0j, 3.39668529 + 0.0j],
[0.95079059 + 0.0j, 7.70464292 + 0.0j],
]
),
np.array(
[
[4.16918831 + 0.0j, 3.4135369 + 0.0j],
[-3.12128344 + 0.0j, 12.22876995 + 0.0j],
]
),
]
expected_variances = [
np.array([[0.00041475, 0.00026418], [0.1143862, 0.07286062]]),
np.array([[0.00432538, 0.00572858], [2.26643718, 3.00169024]]),
]
np.testing.assert_equal(regressor.evalFreq, expected_evalfreq)
np.testing.assert_almost_equal(regressor.parameters[0], expected_impedances[0])
np.testing.assert_almost_equal(regressor.parameters[1], expected_impedances[1])
np.testing.assert_almost_equal(regressor.variances[0], expected_variances[0])
np.testing.assert_almost_equal(regressor.variances[1], expected_variances[1])