ERROR_fun: error function

In jiji6454/Rpac_compReg: Compositional Data Regression

Description

calculate classification, prediction and accuracy errors for fitted FuncompCGL model in simulation.

Usage

ERROR_fun(beta_fit, beta_true, basis_fit, basis_true, sseq, m, p, Nzero_group,
  tol = 0)

Arguments

beta_fit	fitted coefficients vector
beta_true	true coefficientse vector
basis_fit	basis matrix with selected df k and basis_fun
basis_true	basis matrix with true df k and basis_fun
sseq	sequence of time points used to generate basis matrix
m	number of time-invariant variables
p	number of compositional varialbes
Nzero_group	number of true None zero group
tol	tolerance

Value

a list

Non.zero	None zero rows in coefficient matrix beta_C.
class_error	classification errors, including FP, FN, FPR, FNR, etc.
coef_error	L2.cont L2 norm for time-invariant coefficients error L2.comp, L2_inf.comp, L2_L1.comp if selected df k is the same as true degree freedom of basis, coefficients matrix error is calculated
curve_error	norms for curve error by function norms
group_norm	L2 norm for rows in coefficients matrix beta_C

jiji6454/Rpac_compReg documentation built on May 31, 2019, 5:01 a.m.