def testGenerateIndicatorVertices(self):
egoGenerator = EgoGenerator()
numVertices = 500000
means = numpy.array([1, 10])
vars = numpy.array([[5, 1], [1, 2]])
p = 0.1
vList = egoGenerator.generateIndicatorVertices(numVertices, means, vars, p)
X = numpy.zeros((numVertices, means.shape[0]+1))
for i in range(0, numVertices):
X[i, :] = vList.getVertex(i)
(means2, vars2) = Util.computeMeanVar(X)
self.assertTrue((X.astype(numpy.int32) == X).all())
self.assertAlmostEquals(numpy.linalg.norm(means2[0:2] - means), 0, places=1)
self.assertAlmostEquals(numpy.linalg.norm(vars2[0:2][:,0:2] - vars), 0, places=0)
self.assertAlmostEquals(p, means2[2],places=2)
#Try non-symmetric variance matrix
vars = numpy.array([[5, 1], [8, 2]])
self.assertRaises(ValueError, egoGenerator.generateIndicatorVertices, numVertices, means, vars, p)
def checkDistributions():
matFileName = "../../data/EgoAlterTransmissions.mat"
examplesList = ExamplesList.readFromMatFile(matFileName)
numFeatures = examplesList.getDataFieldSize("X", 1)
X = examplesList.getDataField("X")[:, 0:numFeatures/2]
Z = examplesList.getDataField("X")[:, numFeatures/2:numFeatures]
y = examplesList.getDataField("y")
A = Z[y==-1, :]
#Now load directly from the CSV file
#Learn the distribution of the egos
eCsvReader = EgoCsvReader()
egoFileName = "../../data/EgoData.csv"
alterFileName = "../../data/AlterData.csv"
egoQuestionIds = eCsvReader.getEgoQuestionIds()
alterQuestionIds = eCsvReader.getAlterQuestionIds()
(X2, titles) = eCsvReader.readFile(egoFileName, egoQuestionIds)
X2[:, eCsvReader.ageIndex] = eCsvReader.ageToCategories(X2[:, eCsvReader.ageIndex])
(mu, sigma) = Util.computeMeanVar(X)
(mu2, sigma2) = Util.computeMeanVar(X2)
(mu3, sigma3) = Util.computeMeanVar(Z)
(mu4, sigma4) = Util.computeMeanVar(A)
#Seems okay. Next check alters
print(("Mean " + str(mu - mu4)))
print(("Variance " + str(numpy.diag(sigma - sigma4))))
"""
Analysis between the Egos in EgoData.csv and those in EgoAlterTransmissions.mat
reveals that the distributions match closely. The main differences are
in the means and variances in Q44A - D, but this isn't too suprising.
"""
"""
def testGenerateIndicatorVertices2(self):
egoGenerator = EgoGenerator()
numVertices = 500000
means = numpy.array([1, 10])
vars = numpy.array([[5, 1], [1, 2]])
mins = numpy.array([-1000, -1000])
maxs = numpy.array([1000, 1000])
p = 0.1
vList = egoGenerator.generateIndicatorVertices2(numVertices, means, vars, p, mins, maxs)
X = numpy.zeros((numVertices, means.shape[0]+1))
for i in range(0, numVertices):
X[i, :] = vList.getVertex(i)
(means2, vars2) = Util.computeMeanVar(X)
self.assertTrue((X.astype(numpy.int32) == X).all())
self.assertAlmostEquals(numpy.linalg.norm(means2[0:2] - means), 0, places=1)
self.assertAlmostEquals(numpy.linalg.norm(vars2[0:2][:,0:2] - vars), 0, places=0)
self.assertAlmostEquals(p, means2[2],places=2)
self.assertTrue((X[:, 0:2].min(0) >= mins).all())
self.assertTrue((X[:, 0:2].max(0) <= maxs).all())
#Try non-symmetric variance matrix
vars = numpy.array([[5, 1], [8, 2]])
self.assertRaises(ValueError, egoGenerator.generateIndicatorVertices2, numVertices, means, vars, p, mins, maxs)
#Test min > max
vars = numpy.array([[5, 1], [1, 2]])
mins = numpy.array([-2, 6])
maxs = numpy.array([10, 5])
self.assertRaises(ValueError, egoGenerator.generateIndicatorVertices2, numVertices, means, vars, p, mins, maxs)
#Test min == max
numVertices = 1000
vars = numpy.array([[5, 1], [1, 2]])
mins = numpy.array([-2, 5])
maxs = numpy.array([10, 5])
vList = egoGenerator.generateIndicatorVertices2(numVertices, means, vars, p, mins, maxs)
for i in range(0, numVertices):
self.assertTrue(vList.getVertex(i)[1] == 5)
#Try a new example with small range of min and max - check the mean and var
numVertices = 500000
means = numpy.array([1, 10])
vars = numpy.array([[2, 0], [0, 2]])
mins = numpy.array([-3, 6])
maxs = numpy.array([5, 14])
p = 0.1
vList = egoGenerator.generateIndicatorVertices2(numVertices, means, vars, p, mins, maxs)
X = numpy.zeros((numVertices, means.shape[0]+1))
for i in range(0, numVertices):
X[i, :] = vList.getVertex(i)
(means2, vars2) = Util.computeMeanVar(X)
self.assertAlmostEquals(numpy.linalg.norm(means2[0:2] - means), 0, places=1)
self.assertAlmostEquals(numpy.linalg.norm(vars2[0:2][:,0:2] - vars), 0, places=0)
self.assertAlmostEquals(p, means2[2],places=2)
logging.debug((X[:, 0:2].min(0)))
logging.debug((X[:, 0:2].max(0)))
self.assertTrue((X[:, 0:2].min(0) >= mins).all())
self.assertTrue((X[:, 0:2].max(0) <= maxs).all())
