def testEstimateNormal(self):
"""
This passes in a known set of data and ensures the estimateNormal
function returns the expected results.
"""
# 100 samples drawn from mean=0.4, stdev = 0.5
samples = numpy.array([
0.32259025, -0.44936321, -0.15784842, 0.72142628, 0.8794327,
0.06323451, -0.15336159, -0.02261703, 0.04806841, 0.47219226,
0.31102718, 0.57608799, 0.13621071, 0.92446815, 0.1870912,
0.46366935, -0.11359237, 0.66582357, 1.20613048, -0.17735134,
0.20709358, 0.74508479, 0.12450686, -0.15468728, 0.3982757,
0.87924349, 0.86104855, 0.23688469, -0.26018254, 0.10909429,
0.65627481, 0.39238532, 0.77150761, 0.47040352, 0.9676175,
0.42148897, 0.0967786, -0.0087355, 0.84427985, 1.46526018,
1.19214798, 0.16034816, 0.81105554, 0.39150407, 0.93609919,
0.13992161, 0.6494196, 0.83666217, 0.37845278, 0.0368279,
-0.10201944, 0.41144746, 0.28341277, 0.36759426, 0.90439446,
0.05669459, -0.11220214, 0.34616676, 0.49898439, -0.23846184,
1.06400524, 0.72202135, -0.2169164, 1.136582, -0.69576865,
0.48603271, 0.72781008, -0.04749299, 0.15469311, 0.52942518,
0.24816816, 0.3483905, 0.7284215, 0.93774676, 0.07286373,
1.6831539, 0.3851082, 0.0637406, -0.92332861, -0.02066161,
0.93709862, 0.82114131, 0.98631562, 0.05601529, 0.72214694,
0.09667526, 0.3857222, 0.50313998, 0.40775344, -0.69624046,
-0.4448494, 0.99403206, 0.51639049, 0.13951548, 0.23458214,
1.00712699, 0.40939048, -0.06436434, -0.02753677, -0.23017904
])
params = an.estimateNormal(samples)
self.assertWithinEpsilon(params["mean"], 0.3721)
self.assertWithinEpsilon(params["variance"], 0.22294)
self.assertWithinEpsilon(params["stdev"], 0.47216)
self.assertEqual(params["name"], "normal")
def testEstimateNormal(self):
"""
This passes in a known set of data and ensures the estimateNormal
function returns the expected results.
"""
# 100 samples drawn from mean=0.4, stdev = 0.5
samples = numpy.array(
[0.32259025, -0.44936321, -0.15784842, 0.72142628, 0.8794327,
0.06323451, -0.15336159, -0.02261703, 0.04806841, 0.47219226,
0.31102718, 0.57608799, 0.13621071, 0.92446815, 0.1870912,
0.46366935, -0.11359237, 0.66582357, 1.20613048, -0.17735134,
0.20709358, 0.74508479, 0.12450686, -0.15468728, 0.3982757,
0.87924349, 0.86104855, 0.23688469, -0.26018254, 0.10909429,
0.65627481, 0.39238532, 0.77150761, 0.47040352, 0.9676175,
0.42148897, 0.0967786, -0.0087355, 0.84427985, 1.46526018,
1.19214798, 0.16034816, 0.81105554, 0.39150407, 0.93609919,
0.13992161, 0.6494196, 0.83666217, 0.37845278, 0.0368279,
-0.10201944, 0.41144746, 0.28341277, 0.36759426, 0.90439446,
0.05669459, -0.11220214, 0.34616676, 0.49898439, -0.23846184,
1.06400524, 0.72202135, -0.2169164, 1.136582, -0.69576865,
0.48603271, 0.72781008, -0.04749299, 0.15469311, 0.52942518,
0.24816816, 0.3483905, 0.7284215, 0.93774676, 0.07286373,
1.6831539, 0.3851082, 0.0637406, -0.92332861, -0.02066161,
0.93709862, 0.82114131, 0.98631562, 0.05601529, 0.72214694,
0.09667526, 0.3857222, 0.50313998, 0.40775344, -0.69624046,
-0.4448494, 0.99403206, 0.51639049, 0.13951548, 0.23458214,
1.00712699, 0.40939048, -0.06436434, -0.02753677, -0.23017904])
params = an.estimateNormal(samples)
self.assertWithinEpsilon(params["mean"], 0.3721)
self.assertWithinEpsilon(params["variance"], 0.22294)
self.assertWithinEpsilon(params["stdev"], 0.47216)
self.assertEqual(params["name"], "normal")
def testSampleDistribution(self):
"""
Test that sampleDistribution from a generated distribution returns roughly
the same parameters.
"""
# 1000 samples drawn from mean=0.4, stdev = 0.1
p = {"mean": 0.5, "name": "normal", "stdev": math.sqrt(0.1), "variance": 0.1}
samples = _sampleDistribution(p, 1000)
# Ensure estimate is reasonable
np = an.estimateNormal(samples)
self.assertWithinEpsilon(p["mean"], np["mean"], 0.1)
self.assertWithinEpsilon(p["variance"], np["variance"], 0.1)
self.assertWithinEpsilon(p["stdev"], np["stdev"], 0.1)
self.assertTrue(np["name"], "normal")
def testSampleDistribution(self):
"""
Test that sampleDistribution from a generated distribution returns roughly
the same parameters.
"""
# 1000 samples drawn from mean=0.4, stdev = 0.1
p = {"mean": 0.5,
"name": "normal",
"stdev": math.sqrt(0.1),
"variance": 0.1}
samples = _sampleDistribution(p, 1000)
# Ensure estimate is reasonable
np = an.estimateNormal(samples)
self.assertWithinEpsilon(p["mean"], np["mean"], 0.1)
self.assertWithinEpsilon(p["variance"], np["variance"], 0.1)
self.assertWithinEpsilon(p["stdev"], np["stdev"], 0.1)
self.assertTrue(np["name"], "normal")