def testBetaPDF3(self):
print "\n Beta PDF 3: Symmetric"
Points = 6
ZPoints = 101
ZmeanPoints = 1
ZvarPoints = 1
ZMin = 0
ZMax = 1
ZmeanMin = 0.5
ZmeanMax = ZmeanMin
ZvarMin = 0.05
ZvarMax = ZvarMin
Nodes = 10
# create arrays of type "double *"
Z = np.linspace(ZMin, ZMax, ZPoints)
Zmean = np.linspace(ZmeanMin, ZmeanMax, ZmeanPoints)
Zvar = np.linspace(ZvarMin, ZvarMax, ZvarPoints)
# create instances of PDF class
b = pdf.BetaPDF(Zmean, Zvar)
bPdfValM = matrix3d.Matrix3D(ZvarPoints, ZmeanPoints, ZPoints)
bPDF = np.zeros(Points)
# expected PDF values
PDF = np.zeros(ZPoints)
PDF[0] = 0
PDF[1] = 0.96
PDF[2] = 1.44
PDF[3] = 1.44
PDF[4] = 0.96
PDF[5] = 0
# calculate PDF
bTest = b.pdfVal(Z, bPdfValM)
# test
for k in range(Points):
bPDF[k] = bPdfValM.GetVal(0, 0, 20 * k)
"""
print "PDF[0] = inf, bPDF[0] = " + str(bPDF[0])
print "PDF[1] = " + str(PDF[1]) + ", bPDF[1] = " + str(bPDF[1])
print "PDF[2] = " + str(PDF[2]) + ", bPDF[2] = " + str(bPDF[2])
print "PDF[3] = " + str(PDF[3]) + ", bPDF[3] = " + str(bPDF[3])
print "PDF[4] = " + str(PDF[4]) + ", bPDF[4] = " + str(bPDF[4])
print "PDF[5] = inf, bPDF[5] = " + str(bPDF[5])
"""
self.assertTrue(np.abs(bPDF[1] - PDF[1]) < 0.2)
self.assertTrue(np.abs(bPDF[2] - PDF[2]) < 0.2)
self.assertTrue(np.abs(bPDF[3] - PDF[3]) < 0.2)
self.assertTrue(np.abs(bPDF[4] - PDF[4]) < 0.2)
self.assertEqual(bTest, 0)
# create Integrators
Trapz = integrator.Trapz()
Quadr = integrator.GLQuad(Nodes)
# create Filtered Data Matrices
postTrapz = matrix.Matrix(ZvarPoints, ZmeanPoints)
postQuadr = matrix.Matrix(ZvarPoints, ZmeanPoints)
# create matrix for printing
filterTrapz = np.zeros(ZmeanPoints)
filterQuadr = np.zeros(ZmeanPoints)
# create test data
testData = np.ones(ZPoints)
# calculate filtered reaction rates
c = convolute.convVal_func(Z, testData, bPdfValM, postTrapz, Trapz)
c = convolute.convVal_func(Z, testData, bPdfValM, postQuadr, Quadr)
for j in range(ZmeanPoints):
filterTrapz[j] = postTrapz.GetVal(0, j)
filterQuadr[j] = postQuadr.GetVal(0, j)
self.assertTrue(np.abs(filterTrapz[j] - 1) < 0.01)
self.assertTrue(np.abs(filterQuadr[j] - 1) < 0.01)
def testBetaPDF1(self):
print "\n Beta PDF 1: Zero Variance"
ZPoints = 5
ZmeanPoints = 1
ZvarPoints = 1
ZMin = 0
ZMax = 1
ZmeanMin = 0.25
ZmeanMax = ZmeanMin
ZvarMin = 0
ZvarMax = ZvarMin
Nodes = 50
# create arrays of type "double *"
Z = np.linspace(ZMin, ZMax, ZPoints)
Zmean = np.linspace(ZmeanMin, ZmeanMax, ZmeanPoints)
Zvar = np.linspace(ZvarMin, ZvarMax, ZvarPoints)
# create instances of PDF class
b = pdf.BetaPDF(Zmean, Zvar)
bPdfValM = matrix3d.Matrix3D(ZvarPoints, ZmeanPoints, ZPoints)
# expected PDF values
PDF = np.zeros(ZPoints)
PDF[1] = 1 * (ZPoints - 1)
# calculate PDF
test = b.pdfVal(Z, bPdfValM)
PDFCalc = np.zeros(ZPoints)
# test
for k in range(ZPoints):
PDFCalc[k] = bPdfValM.GetVal(0, 0, k)
self.assertAlmostEqual(PDF[k], PDFCalc[k])
# create Integrators
Trapz = integrator.Trapz()
Quadr = integrator.GLQuad(Nodes)
# create Filtered Data Matrices
postTrapz = matrix.Matrix(ZvarPoints, ZmeanPoints)
postQuadr = matrix.Matrix(ZvarPoints, ZmeanPoints)
# create matrix for printing
filterTrapz = np.zeros(ZmeanPoints)
filterQuadr = np.zeros(ZmeanPoints)
# create test data
testData = np.ones(ZPoints)
# calculate filtered reaction rates
c = convolute.convVal_func(Z, testData, bPdfValM, postTrapz, Trapz)
c = convolute.convVal_func(Z, testData, bPdfValM, postQuadr, Quadr)
for j in range(ZmeanPoints):
filterTrapz[j] = postTrapz.GetVal(0, j)
filterQuadr[j] = postQuadr.GetVal(0, j)
self.assertEqual(filterTrapz[j], 1)
self.assertAlmostEqual(filterQuadr[j], 1, 1)
if locflag == 0:
raise IOError("Production rate data for %s is missing" %
speciesprodrate)
# Obtain relevant Yi and reaction rates from each file, and convolute
integ = options["Integrator"][0]
if integ == 'trapezoid':
Intgr = integrator.Trapz()
elif integ == 'simpson':
Intgr = integrator.Simpson()
elif integ == 'glquad':
NumberNodes = iof.read_input("glq Number of Nodes:",
inputs,
minargs=0,
default=[50])
Intgr = integrator.GLQuad(int(NumberNodes[0]))
else:
raise IOError(
"inavlid integrator type (%s) specified, instead use <trapezoid>, <simpson>, or <glquad>"
% integ)
print "Convoluting using %s integration" % integ
convolutedC = [0] * nofiles
convolutedST = [0] * nofiles
maxC = 0
maxRate = 0
for kk in range(nofiles): # future verisons: add loop over [C ST Y1 Y2 etc]
file = datafiles[int(filesmatC.GetVal(kk, 1))]
massfracs = np.genfromtxt(file,
unpack=False,
skip_header=2,