R/ex_select_bw_tmp.R

In gbasulto/kerdec: Kernel deconvolution methods

# ## Case 1: normal error with known variance (misspecified
# ## distribution)
# case1 <-
#   kerdec_dens(Y, method = "CV", kernel = "sinc",
#               lower = lower, upper = upper,
#               error_dist = "laplace",
#               error_scale_par = sd_error,
#               # bw_interval = c(0.08, 0.5),
#               resolution = 128)
# 
# aaa <- select_bw(Y, "CV", "flat", error_smp = eps,
#                  error_dist = "laplace",
#                  bw_interval = c(0.08, 0.5))
# aaa
# 
# with(case1, plot(x_eval, f_vals, type = "l"))
# 
# case1 <-
#   kerdec_dens(Y, method = "CV", kernel = "vp",
#               lower = lower, upper = upper,
#               error_dist = "normal",
#               error_scale_par = sd_error,
# #              bw_interval = c(0.08, 0.5),
#               resolution = 128)
# 
# with(case1, plot(x_eval, f_vals, type = "l"))
# 
# case1 <-
#   kerdec_dens(Y, method = "CV", kernel = "triw",
#               lower = lower, upper = upper,
#               error_dist = "normal",
#               error_scale_par = sd_error,
#               # h0 = 0.19,
#               bw_interval = c(0.07, 0.2),
#               resolution = 128)
# 
# with(case1, plot(x_eval, f_vals, type = "l"))
# 
# case1 <-
#   kerdec_dens(Y, method = "CV", kernel = "tric",
#               lower = lower, upper = upper,
#               error_dist = "normal",
#               error_scale_par = sd_error,
#               # bw_interval = c(0.08, 0.5), 
#               resolution = 128)
# 
# with(case1, plot(x_eval, f_vals, type = "l"))
# 
# case1 <-
#   kerdec_dens(Y, method = "CV", kernel = "flat",
#               lower = lower, upper = upper, h = NULL,
#               error_dist = "normal",
#               error_scale_par = sd_error,
# #              bw_interval = c(0.08, 0.5), 
#               resolution = 128)
# 
# with(case1, plot(x_eval, f_vals, type = "l"))
#