inst/histogram.segment.example.R

In helen-zhu/ConsensusPeaks: A Method for Peak Merging And Distribution Fitting Across Biological Replicates

library(ConsensusPeaks)

set.seed(13)
# Example 1, uniform
x = obs.to.int.hist(runif(1000, min = 0, max = 50))
s = ftc(x)
plot.segments(x, s)

#set.seed(13)
# set.seed(26)
# Example 2, normal dist
x.norm = obs.to.int.hist(rnorm(1000) * 5)
s.norm = ftc(x.norm)
plot.segments(x.norm, s.norm)

set.seed(26)
# Example 2.1, normal dist
x.norm = obs.to.int.hist(rnorm(1000) * 5)
s.norm = ftc(x.norm)
plot.segments(x.norm, s.norm)

# Example 1, uniform
s = ftc.helen(x)
plot.segments(x, s)

#set.seed(13)
# set.seed(26)
# Example 2, normal dist
x.norm = obs.to.int.hist(rnorm(1000) * 5)
s.norm = ftc.helen(x.norm)
plot.segments(x.norm, s.norm)

set.seed(26)
# Example 2.1, normal dist
x.norm = obs.to.int.hist(rnorm(1000) * 5)
s.norm.helen = ftc.helen(x.norm, s = local.minmax(x.norm))
s.norm = ftc(x.norm)
plot.segments(x.norm, unlist(s.norm.helen), main = "x.norm Helen's FTC")
plot.segments(x.norm, s.norm, main = "x.norm FTC")


# Example 2, messy mixture of gaussians and uniform
set.seed(26)
# x.norm.mix = obs.to.int.hist(c(rnorm(25, mean = 1), rnorm(10, mean = 5), c(runif(20, min = 3, max = 5))))
x.norm.mix = obs.to.int.hist(c(rnorm(25, mean = 1), rnorm(80, mean = 10)))
s.norm.mix.helen = ftc.helen(x.norm.mix, s = local.minmax(x.norm.mix))
s.norm.mix = ftc(x.norm.mix)
plot.segments(x.norm.mix, unlist(s.norm.mix.helen), main = "x.norm.mix Helen's FTC")
plot.segments(x.norm.mix, s.norm.mix, main = "x.norm.mix FTC")