def runTest(self):
luckCalcPlot = LuckyCalculations(self.dsData,
self.calib,
self.integConf,
self.bulbTemp,
"TestData",
debug=False)
luckCalcPlot.runCalculations()
LuckyPlots(luckCalcPlot)
print "Test sleeping for 20s"
time.sleep(20)
'''
Created on 12 Jan 2016
@author: wnm24546
'''
import os
import numpy as np
from Lucky.Calculations import LuckyCalculations
wdir = os.path.join('.','testData')
dsDataFile = os.path.join(wdir, 'T_635.txt')
usDataFile = os.path.join(wdir, 'T_636.txt')
dsCalibFile = os.path.join(wdir, 'Calib.txt')
usCalibFile = os.path.join(wdir, 'CalibF2MTW.txt')
dsData = np.loadtxt(dsDataFile, unpack=True) ##Raw file
usData = np.loadtxt(usDataFile, unpack=True)
calib = np.loadtxt('./testData/Calib.txt', unpack=True) ##Calib file
integConf = [315, 800, 200] #Values lifted out of PreLucky_Variant.py
bulbTemp = 2436
luckCalc = LuckyCalculations(dsData, calib, integConf, bulbTemp, "TestData", debug=True)
luckCalc.runCalculations()
print calib
class CalculationsTest(LuckyCalculationsTest):
def setUp(self):
LuckyCalculationsTest.setUp(self)
self.luckCalc = LuckyCalculations(self.dsData, self.calib, self.integConf, self.bulbTemp, "TestData", debug=True)
self.luckCalc.runCalculations()
self.workingCalcs(self.dsData, self.calib, self.integConf)
def workingCalcs(self, data, calib, integConf):
from scipy.constants import h, c, k, pi
##Constants:
Kb = k
##Session where I define all the function needed
def Planck(x,e,T):
x = x*10**(-9)
a=np.expm1((h*c)/(k*x*T))#Order changed!! (h*c/k)/(x)/T
P=e/(x)**5*(2*pi*h*c**2)*1/(a) ####NB Removed
return P
#Defined Wien function
def Wien(Int,x):
#Order changed!!
#W=Kb/h/c*np.log((x*10**(-9))**5*Int/2/pi/h/c**2)
W=Kb/(h*c)*np.log((x*10**(-9))**5*Int/(2*pi*h*c**2))
return W
#Defined two-colour function
def TwoCol(Int,x):
count=len(x)
delta=200
k=count-delta
TTwo=[]*count
for i in range (0,k):#(0,count-1):
f1=(h*c)/(x[i]*10**(-9)*Kb)
f2=(h*c)/(x[i+delta]*10**(-9)*Kb)
i1=np.log(Int[i]*(x[i]*10**(-9))**5/(2*pi*h*c**2))*(Kb/(h*c))#Order changed!!
i2=np.log(Int[i+delta]*(x[i+delta]*10**(-9))**5/(2*pi*h*c**2))*(Kb/(h*c))#i2=np.log(Int[i]/2/pi/h/c**2/f1**5)*Kb/h/c
TTwo.append((f1-f2)/(i2-i1))
for i in range (k,count):
a = float('nan')
TTwo.append(a)
return TTwo
#Defined linear fit for Wien function
def FWien(x,e,T):
a=1/T
b=Kb/(h*c)*np.log(e)
W=b-a*x
return W
#Defined Gauss fit
def gaus(x, a, x0, sigma):
return np.real(a*np.exp(-(x-x0)**2/(2*sigma**2)))
self.x = data[0]
y = data[1]
xC = calib[0]
yC = calib[1]
start = integConf[0]
end = integConf[1]
delta = integConf[2]
P=Planck(self.x,1,2436)##Ideal Planck
self.P = np.reshape(P, (1, len(P)))
self.Norm=y/yC*P #Normalization file
self.invX=1e9/self.x #Inverse of wavelength for Wien function (CHANGED 1/x*10**9)
self.W=Wien(self.Norm,self.x)
self.Two=TwoCol(self.Norm,self.x)
Two2=np.array(self.Two,dtype='float')
self.TwoInt=Two2[start:end]
bins=range(1000,3000,1)
hist=np.histogram(self.TwoInt,bins,density=False)
self.freq=np.array(hist[0])
control=len(hist[1])-1
self.value=np.array(np.delete(hist[1],control,0))
p0=[1,2000]
#Fit Planck in the range [start:end]
self.xp=self.x[start:end]
self.Normp=self.Norm[start:end]
bestP,covarP = curve_fit(Planck, self.xp, self.Normp, p0)
TP=round(bestP[1],2)
self.TPNR = bestP[1]
self.eP=bestP[0]#Save planck Emissivity
FP=Planck(self.xp,self.eP,TP)#Create the new Planck with the fit parameters
self.FPNR = Planck(self.xp,self.eP,self.TPNR)
PRes=abs(self.Normp-FP)#Planck Residual
#Wien fit
self.invX1=self.invX[start:end]
self.W1=self.W[start:end]
#Fit Wien and control that there are no inf or nan arguments in the fit
self.bestW,covarW = curve_fit(FWien,self.invX1[(np.isfinite(self.W1))],self.W1[(np.isfinite(self.W1))],p0=[1,self.TPNR]) #Changed - was TP
self.Residual=self.W1-FWien(self.invX1[(np.isfinite(self.W1))],*self.bestW)
#Save Wien temperature
TW=round(self.bestW[1])
#Gaussian fit to the histogram two-colours
popt,pcov = curve_fit(gaus,self.value,self.freq,p0=[1000,self.TPNR,100])#Changed - was TP
Thist=round(popt[1],2)#Save Histogram temperature
errTot=round(popt[2])
self.Thist = popt[1]
self.errTot = popt[2]
class CalculationsTest(LuckyCalculationsTest):
def setUp(self):
LuckyCalculationsTest.setUp(self)
self.luckCalc = LuckyCalculations(self.dsData,
self.calib,
self.integConf,
self.bulbTemp,
"TestData",
debug=True)
self.luckCalc.runCalculations()
self.workingCalcs(self.dsData, self.calib, self.integConf)
def workingCalcs(self, data, calib, integConf):
from scipy.constants import h, c, k, pi
##Constants:
Kb = k
##Session where I define all the function needed
def Planck(x, e, T):
x = x * 10**(-9)
a = np.expm1((h * c) / (k * x * T)) #Order changed!! (h*c/k)/(x)/T
P = e / (x)**5 * (2 * pi * h * c**2) * 1 / (a) ####NB Removed
return P
#Defined Wien function
def Wien(Int, x):
#Order changed!!
#W=Kb/h/c*np.log((x*10**(-9))**5*Int/2/pi/h/c**2)
W = Kb / (h * c) * np.log(
(x * 10**(-9))**5 * Int / (2 * pi * h * c**2))
return W
#Defined two-colour function
def TwoCol(Int, x):
count = len(x)
delta = 200
k = count - delta
TTwo = [] * count
for i in range(0, k): #(0,count-1):
f1 = (h * c) / (x[i] * 10**(-9) * Kb)
f2 = (h * c) / (x[i + delta] * 10**(-9) * Kb)
i1 = np.log(Int[i] * (x[i] * 10**(-9))**5 /
(2 * pi * h * c**2)) * (Kb /
(h * c)) #Order changed!!
i2 = np.log(Int[i + delta] * (x[i + delta] * 10**(-9))**5 /
(2 * pi * h * c**2)) * (
Kb / (h * c)
) #i2=np.log(Int[i]/2/pi/h/c**2/f1**5)*Kb/h/c
TTwo.append((f1 - f2) / (i2 - i1))
for i in range(k, count):
a = float('nan')
TTwo.append(a)
return TTwo
#Defined linear fit for Wien function
def FWien(x, e, T):
a = 1 / T
b = Kb / (h * c) * np.log(e)
W = b - a * x
return W
#Defined Gauss fit
def gaus(x, a, x0, sigma):
return np.real(a * np.exp(-(x - x0)**2 / (2 * sigma**2)))
self.x = data[0]
y = data[1]
xC = calib[0]
yC = calib[1]
start = integConf[0]
end = integConf[1]
delta = integConf[2]
P = Planck(self.x, 1, 2436) ##Ideal Planck
self.P = np.reshape(P, (1, len(P)))
self.Norm = y / yC * P #Normalization file
self.invX = 1e9 / self.x #Inverse of wavelength for Wien function (CHANGED 1/x*10**9)
self.W = Wien(self.Norm, self.x)
self.Two = TwoCol(self.Norm, self.x)
Two2 = np.array(self.Two, dtype='float')
self.TwoInt = Two2[start:end]
bins = range(1000, 3000, 1)
hist = np.histogram(self.TwoInt, bins, density=False)
self.freq = np.array(hist[0])
control = len(hist[1]) - 1
self.value = np.array(np.delete(hist[1], control, 0))
p0 = [1, 2000]
#Fit Planck in the range [start:end]
self.xp = self.x[start:end]
self.Normp = self.Norm[start:end]
bestP, covarP = curve_fit(Planck, self.xp, self.Normp, p0)
TP = round(bestP[1], 2)
self.TPNR = bestP[1]
self.eP = bestP[0] #Save planck Emissivity
FP = Planck(self.xp, self.eP,
TP) #Create the new Planck with the fit parameters
self.FPNR = Planck(self.xp, self.eP, self.TPNR)
PRes = abs(self.Normp - FP) #Planck Residual
#Wien fit
self.invX1 = self.invX[start:end]
self.W1 = self.W[start:end]
#Fit Wien and control that there are no inf or nan arguments in the fit
self.bestW, covarW = curve_fit(FWien,
self.invX1[(np.isfinite(self.W1))],
self.W1[(np.isfinite(self.W1))],
p0=[1, self.TPNR]) #Changed - was TP
self.Residual = self.W1 - FWien(self.invX1[(np.isfinite(self.W1))], *
self.bestW)
#Save Wien temperature
TW = round(self.bestW[1])
#Gaussian fit to the histogram two-colours
popt, pcov = curve_fit(gaus,
self.value,
self.freq,
p0=[1000, self.TPNR, 100]) #Changed - was TP
Thist = round(popt[1], 2) #Save Histogram temperature
errTot = round(popt[2])
self.Thist = popt[1]
self.errTot = popt[2]
def runTest(self):
luckCalcPlot = LuckyCalculations(self.dsData, self.calib, self.integConf, self.bulbTemp, "TestData", debug=False)
luckCalcPlot.runCalculations()
LuckyPlots(luckCalcPlot)
print "Test sleeping for 20s"
time.sleep(20)