Description Usage Arguments Value Examples
View source: R/Fitting_algorithm.R
Main fitting algorithm of FSPDM method
1 2 3 | train_function(Data_generated, Eig_num, k, beta, theta, sigma2,
lambda1 = 0, lambda2 = 0, lambda3 = 0, lambda4 = 0,
testIndexes = NULL, sigma2_list = NULL, maxout = 1)
|
Data_generated |
a list constaining the observed value of functional data, covariates and penalty matrix |
Eig_num |
the number of eigenfunction users want to use |
k |
dimension of spline basis with respect to covariate in eigenfunctions |
beta |
the start point of beta parameters, can be set with all zeros |
theta |
the start point of theta parameters, can be set with all zeros |
sigma2 |
the start point of sigma2 parameters, can be set with small positive value |
lambda1 |
the tuning parameters of smoothness penalty with respect to t in mean function |
lambda2 |
the tuning parameters of smoothness penalty with respect to covariate in mean function |
lambda3 |
the tuning parameters of smoothness penalty with respect to t in eigenfunction |
lambda4 |
he tuning parameters of smoothness penalty with respect to covariate in eigenfunction |
testIndexes |
test Index if use cv |
sigma2_list |
measurement error if have |
maxout |
algorithm setting, default 1 |
a list constains all estimator of SFPDM method
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 34 | tmin = 0 # the start point of the curve
tmax = 1 # the end point of the curve, i.e. the cuvre's regin is from tmin to tmax
ymin = 0 # the minimum of the covariates
ymax = 1 # the maximum of the covariate
num_bin = 20 # number of bin in initial steps of our algorithm
order = 4 # spline order
nknots =8 # number of knots
splineObj_t = new(orthoSpline,tmin,tmax,order,nknots)
# degree freedom of spline basis
M1 = splineObj_t$getDoF()
## basis with respect to y to get the tensor product basis
yknots = 3
splineObj_y = new(orthoSpline,ymin,ymax,order,(yknots))
# degree freedom of spline basis
M2 = splineObj_y$getDoF()
## basis with d that is in matrix C
# basis with respect y (d^T), in C matrix
dknots = 5
splineObj_d = new(orthoSpline,ymin,ymax,order,dknots)
# degree freedom of spline basis
M3 = splineObj_d$getDoF()
k = M3
Spline_func = c(splineObj_t,splineObj_d,splineObj_y)
## set the number of principle components
Eig_num = r = 3
## the the number of spline basis used
## spline basis for t
M1 = 10
M2 = 5
M3 = k = 7
theta = rep(0,M1*M2)
sigma2 = 0.01
beta = rep(0,r*M1*M3)
parameter_best = train_function (Data_generated = Data_generated,Eig_num = Eig_num,k = k, beta = beta,theta = theta,sigma2 = sigma2)
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.