corinf: Inf-correction of cummulative weights of a sample
In RostyslavMaiboroda/mixvconc:

Description Usage Arguments Details Value See Also Examples

corinf calculates cummulative weights for the lower nondecreasing envelope of the CDFs of mixture components (inf-corrected weights). The weights are truncated by 0.

1	corinf(xs)

`xs`	object from class `wtsamp` containing the sample and weights for components' distributions.

If cummulative weights are NULL they will be calculated from the individual ones by indiv2cum.

object from class wtsamp whose cummulative weights are inf-corrected and individual weights are set to NULL.

Maiboroda R. and Kubaichuk O. Asymptotic normality of improved weighted empirical distribution functions. Theor. Probability and Math. Statist. 69 (2004), 95-102

p <- genunifp(15,2) # create mixing probabilities
a <- lsweight(p) # calculate minimax weights
# create a weighted sample:
xs <- wtsamp(genormixt(p,c(0,1),c(1,1)),indiv=a)
F1 <- edfgen(xs,1) # minimax the estimate for 1-st component
xs_inf <- corinf(xs)
F1_inf <- edfgen(xs_inf,1) # inf-corrected estimate for 1-st component
F1(0)
F1_inf(0)
plot(F1,-3,3)
curve(F1_inf,col="red",,lty="dashed",add=TRUE)