def test_fit_with_fixed_parameter(self):
"""
Linear fit for y=ax+b where a is fixed.
"""
# Create invariant object. Background and scale left as defaults.
fit = invariant.Extrapolator(data=self.data)
p, dp = fit.fit(power=-1.0)
# Test results
self.assertAlmostEqual(p[0], 1.0, 5)
self.assertAlmostEqual(p[1], 0.0, 5)
def test_fit_linear_data(self):
"""
Simple linear fit
"""
# Create invariant object. Background and scale left as defaults.
fit = invariant.Extrapolator(data=self.data)
#a,b = fit.fit()
p, dp = fit.fit()
# Test results
self.assertAlmostEqual(p[0], 1.0, 5)
self.assertAlmostEqual(p[1], 0.0, 5)
def test_fit_line_data_fixed_no_weight(self):
"""
Fit_Test_2: test linear fit, ax +b, with 'a' fixed
"""
# Create invariant object. Background and scale left as defaults.
fit = invariant.Extrapolator(data=self.data)
# Fixing a = -power =4
p, dp = fit.fit(power=-4)
# Test results
self.assertAlmostEqual(p[0], 4)
self.assertAlmostEqual(p[1], -7.8, 3)
def skip_test_fit_line_data_no_weight(self):
"""
Fit_Test_1: test linear fit, ax +b, without fixing the power a
"""
# Create invariant object. Background and scale left as defaults.
fit = invariant.Extrapolator(data=self.data)
# Let the power float (not fixed)
p, dp = fit.fit(power=None)
# Test results
self.assertAlmostEqual(p[0], 2.4727, 3)
self.assertAlmostEqual(p[1], 0.6, 3)
def test_fit_line_data(self):
"""
Fit_Test_1: test linear fit, ax +b, without fixed
"""
# Create invariant object. Background and scale left as defaults.
fit = invariant.Extrapolator(data=self.data)
# Without holding
p, dp = fit.fit(power=None)
# Test results
self.assertAlmostEquals(p[0], 2.3983, 3)
self.assertAlmostEquals(p[1], 0.87833, 3)
def test_fit_linear_data_with_noise_and_fixed_par(self):
"""
Simple linear fit with noise
"""
import random, math
for i in range(len(self.data.y)):
self.data.y[i] = self.data.y[i]+.1*(random.random()-0.5)
# Create invariant object. Background and scale left as defaults.
fit = invariant.Extrapolator(data=self.data)
p, dp = fit.fit(power=-1.0)
# Test results
self.assertTrue(math.fabs(p[0]-1.0)<0.05)
self.assertTrue(math.fabs(p[1])<0.1)
def test_fit_line_data(self):
"""
Fit_Test_1: test linear fit, ax +b. The power law is calculated
on the log data so that B * x^A becomes logB + A logX = ax+b.
For the extrapolation a can be fixed (usually to -4) or floating.
Here we test letting it float.
"""
# Create invariant object. Background and scale left as defaults.
fit = invariant.Extrapolator(data=self.data)
# Let the power float (not fixed)
p, dp = fit.fit(power=None)
# Test results
self.assertAlmostEqual(p[0], 2.3983, 3)
self.assertAlmostEqual(p[1], 0.87833, 3)
