L2_deconvolution: L2 deconconvolution function

In hernanmp/RRDecon:

Description

This function allows to perform L2 deconvolution with model selection as in the paper "A deconvolution path for mixtures"

Usage

L2_deconvolution(y, prop = 0.25, d = floor((length(y)^(1/(2.01)))),
  lambda_grid = sort(c(0.001, 0.1, 5, seq(10, 400, length = 16), seq(500,
  2000, length = 10), 20000)))

Arguments

y	: a vector containing the raw observations.
prop	: a scalar between 0 and 1 indicating the proportion of samples to be used as held out set. The default choice is 0.25
d	: number of bins, the default choice is floor( (length(y)^(1/(2.01)) ))
lambda_grid	: list of reuglarization parameters, default choice is sort(c(0.001,0.1,5,seq(10,400,length = 16),seq(500,2000,length = 10),20000))

Value

loc : location of the centers of bins where the density is estimated

f_hat : estimated mixing density

y_hat : marginal density

hernanmp/RRDecon documentation built on May 28, 2019, 8:44 a.m.