test_example_1 added.

Author: wollmich

tests/test_example_1.py

@@ -0,0 +1,48 @@
+# B_LEAST ISO 6143:2001
+# Michael Wollensack METAS - 24.10.2024 - 29.11.2024
+
+# pylint: disable=missing-module-docstring
+# pylint: disable=missing-class-docstring
+# pylint: disable=missing-function-docstring
+# pylint: disable=wildcard-import
+# pylint: disable=unused-wildcard-import
+
+from pytest import approx
+from metas_b_least import *
+
+def test_example_1():
+    # See appendix B.2.1 Example 1, pages 23 - 25
+    print('Example B LEAST 1\n')
+    cal_data = b_read_cal_data(os.path.join(data_dir, 'b_least_1_data_cal.txt'))
+    meas_data = b_read_meas_data(os.path.join(data_dir, 'b_least_1_data_meas.txt'))
+    print('Linear function\n')
+    b, b_cov, b_res, x, x_cov = b_test(cal_data, meas_data, b_linear_func)
+    # intercept b0
+    assert b[0] == approx(-3.5747e-1, abs=0.0001e-1)
+    # slope b1
+    assert b[1] == approx(2.4612e1, abs=0.0001e1)
+    # standard uncertainty of the intercept
+    assert np.sqrt(b_cov[0, 0]) == approx(1.5716e-1, abs=0.0005e-1) # 1.5711e-1
+    # standard uncertainty of the slope
+    assert np.sqrt(b_cov[1, 1]) == approx(4.8048e-1, abs=0.0022e-1) # 4.8026e-1
+    # covariance between intercept and slope
+    assert b_cov[0, 1] == approx(-5.6921e-2, abs=0.00520e-2) # 5.6869e-2
+    # residual
+    assert np.sum(b_res*b_res) == approx(0.6743, abs=0.0001)
+    # goodness of fit
+    assert np.max(np.abs(b_res)) == approx(0.568, abs=0.001)
+    # mixture no 1
+    assert x[0] == approx(5.9923, abs=0.0001)
+    assert np.sqrt(x_cov[0, 0]) == approx(1.6377e-1, abs=0.0001e-1)
+    # mixture no 2
+    assert x[1] == approx(1.4409e1, abs=0.0001e1)
+    assert np.sqrt(x_cov[1, 1]) == approx(3.5599e-1, abs=0.0005e-1) # 3.5594e-1
+    # mixture no 3
+    assert x[2] == approx(4.3943e1, abs=0.0001e1)
+    assert np.sqrt(x_cov[2, 2]) == approx(1.1631, abs=0.0003) # 1.1628
+    # covariance between values for mixtures 1 and 2
+    assert x_cov[0, 1] == approx(1.16e-2, abs=0.01e-2)
+    # covariance between values for mixtures 1 and 3
+    assert x_cov[0, 2] == approx(1.48e-2, abs=0.01e-2)
+    # covariance between values for mixtures 2 and 3
+    assert x_cov[1, 2] == approx(1.37e-1, abs=0.01e-1)