count_channels: Count channels

In jschap1/WSEw: River Geometry Package for Studying Water Surface Elevation-Width Relationships

Description

Returns the number of channels contained within a (possibly multi-channel) cross section

Usage

count_channels(x, d)

Arguments

x	x coordinate
b	bed elevation

Details

Finds the local minima in the channel cross section Uses smoothing techniques described at https://rpubs.com/mengxu/peak_detection Accuracy depends on choice of tuning parameters Could calibrate to a particular dataset

Examples

k <- 1000
x <- cross_sections$x[[k]]
b <- cross_sections$b[[k]]
d <- cross_sections$d[[k]]
plot(x,b,type="l")

Here is some code I've been playing with:
pool <- 21
i <- 8
d <- readRDS(paste0("./Outputs/p", pool, "/main_channel.rds"))
wbf <- readRDS(paste0("./Outputs/p", pool, "/wbf.rds"))
cross_sections <- build_cross_sections(d, refWSE[i], wbf, option = "zero")
head(which(cross_sections$channel.pix>100))
na.ind <- get_na_ind(cross_sections)
head(which(!na.ind), 100)
x <- 4041
plot(cross_sections$x[[x]], cross_sections$b[[x]], type = "l")
count_channels(cross_sections$x[[x]], cross_sections$d[[x]])

w1 <- round(0.2*length(cross_sections$x[[x]])) #' one idea for the w parameter's value
100/wbf[x]

#' Make an animation of how the river channel changes from upstream to downstream
na.ind <- get_na_ind(cross_sections)
for (x in which(!na.ind))
{
  num <- sprintf("%05d", x) #' ensuring the same number of digits so they files can be sorted in the proper order
  png(paste0("pool21_xs_", num, ".png"))
  plot(cross_sections$x[[x]], 
       cross_sections$b[[x]], 
       type = "l", 
       main = paste("xs", x, "wbf = ", round(wbf[x]))
  )
  dev.off()
  #' readline(prompt="Press [enter] to continue")
}

nc <- vector(length = length(which(!na.ind)))
for (x in which(!na.ind))
{
  nc[x] <- count_channels(cross_sections$x[[x]], cross_sections$d[[x]])
  if (x>2000){break}
}

jschap1/WSEw documentation built on May 12, 2019, 10:50 a.m.