Identify data blocks - r

Question

I have a vector with either a negative value or NA and a threshold:
threshold <- -1
example <- c(NA, NA, -0.108, NA, NA, NA, NA, NA -0.601, -0.889, -1.178, -1.089, -1.401, -1.178, -0.959, -1.085, -1.483, -0.891, -0.817, -0.095, -1.305, NA, NA, NA, NA, -0.981, -0.457, -0.003, -0.358, NA, NA)
I want to identify all the data blocks with at least one value lower than the threshold and to replace by NA all the other blocks. With my example vector, I want this result:
result <- c(NA, NA, NA, NA, NA, NA, NA, NA -0.601, -0.889, -1.178, -1.089, -1.401, -1.178, -0.959, -1.085, -1.483, -0.891, -0.817, -0.095, -1.305, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA)
So the first available value is the first block but -0.108 is higher than -1 so it turns into NA. The second block is kept the same because there is at least ine value lower than -1. The third block is now NA values because between the 4 available values, no one was lower than the threshold.
My first idea was to identify where were the values lower than the threshold:
val <- which(example < threshold)
But then I don't know how to say "keep all the values around this position which are not NA" because it is always a different number of values...

Answer 1

Try
library(data.table)#v >= 1.9.5 (devel version - install from GitHub).
#library(devtools)
#install_github("Rdatatable/data.table", build_vignettes = FALSE)
as.data.table(example)[, res:=(NA | (min(example)< -1))*example, by=rleid(is.na(example))][, res]

Answer 2

Another way, with the suggestion of OlliJ :
example <- c(NA, NA, -0.108, NA, NA, NA, NA, NA -0.601, -0.889, -1.178, -1.089, -1.401, -1.178, -0.959, -1.085, -1.483, -0.891, -0.817, -0.095, NA, NA, NA, NA, -0.981, -0.457, -0.003, -0.358, NA, NA)
test <- !(is.na(example))
len <- rle(test)$lengths
val <- rle(test)$values
##Matrix with the beginning and the end of each group
ind <- matrix(,nrow=length(which(val)),ncol=2)
ind[,1] <- (cumsum(len)[which(val==T)-1])+1
ind[,2] <- (cumsum(len))[val==T]
result <- rep(NA, length=length(example))
apply(ind, 1, function(x)
{
if(any(example[x[1]:x[2]] < -1))
{
result[x[1]:x[2]] <- example[x[1]:x[2]]
}
})