mixture_generator(n = 130, p = 100, ratio = 0.4, max_compl = 1,
valid = 1000, positive = 0.6, sigma_Y = 10, sigma_X = NULL,
R2 = NULL, R2Y = 0.4, meanvar = NULL, sigmavar = NULL, lambda = 3,
Amax = NULL, lambdapois = 10, gamma = FALSE, gammashape = 1,
gammascale = 0.5, tp1 = 1, tp2 = 1, tp3 = 1, nonlin = 0,
pnonlin = 2, scale = TRUE, Z = NULL)
gamma=TRUE
or if R2
is not NULLsigma_X
gamma-distributedratio
and max_compl
parameters.p
covariates following Gaussian Mixtures with some of them generated by sub-regressions on others.n
) generated by linear regression on X_appr
with coefficients A
and residual standard deviation sigma_Y
.Y_appr
B[i-1,j]
is the coefficient associated to X_appr[,i]
in the sub-regression that generates X_appr[,j]
p
that describes the structure of sub-regressions. Z[i,j]
=1 if X_appr[,i]
explains X_appr[,j]
X_appr
, with valid
individuals.Y_appr
and Y_test
.require(CorReg)
#dataset generation
base=mixture_generator(n=1500,p=10,valid=0)
X_appr=base$X_appr #learning sample
Y_appr=base$Y_appr#response variable
for(i in 1:ncol(X_appr)){
hist(X_appr[,i])
}
Run the code above in your browser using DataLab