optimalSampling/optimalBatch5.py at main · cossorzano/optimalSampling · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33

import numpy as np
import optimalSampling

class LinearFunction2(optimalSampling.FittingFunctionLS):
    def getFunction(self, xn, beta):
        return beta[0]+beta[1]*xn+beta[2]*xn*xn

    def getPartialDerivative1(self, xn, yn, beta, i):
        if i==0:
            return 1
        elif i==1:
            return xn
        elif i==2:
            return xn*xn

h=LinearFunction2()
h.sigma2=0.1
trueBeta=np.asarray([-0*0,-2*0,1])

X=np.asarray([[0],[0.5],[1]])
y=h.simulateFunctionAtMultiplePoints(X, trueBeta, True)

stepx=0.01

def CramerRaoBound2(I):
    return optimalSampling.CramerRaoBound(I,2)

evaluator0=optimalSampling.FIMEvaluator()
evaluator1=optimalSampling.FIMEvaluator(CramerRaoBound2)
evaluator2=optimalSampling.VarEvaluator()

optimalSampling.simulateProcess(h,trueBeta,X,y,np.asarray([0,0,0]),100,np.mgrid[0:1+stepx:stepx],evaluator0, verbose=True)