brokenstick: Fitting brokenstick models
In SESman/rbl: Biologging tools for diving predators in R

Description Usage Arguments Details Value See Also Examples

brokenstick is used to fit brokenstick models on two-dimensional data such as Time-Depth, Depth-Temperature or Depth-Light profiles. Brokenstick models are useful for compressing or extracting the shape of high resolution profiles.

1 2	brokenstick(x, y, npts = 6, start = NULL, na.action, allow.dup = FALSE, not.inj.action = c("ignore", "null", "error"), sort.data = FALSE, ...)

`x`	The x data. Note that the `x` values have to be sorted. Can also be a list of `x` and `y` (processed by `xy.coords`). Alternatively, it can also be an object of class "`formula`".
`y`	The y data. (processed by `xy.coords`). Alternatively, if `x` is a `'formula'`, `y` can also be an optional data frame, list or environment (or object coercible by `as.data.frame` to a data frame) containing the variables in the model. If not found in data, the variables are taken from environment (`formula`), typically the environment from which `brokenstick` is called.
`npts`	The number of points for the brokenstick model to fit. See `optBrokenstick` for a version of `brokenstick` which figures out this number of points automatically.
`start`	Some starting points to start the algorithm with.
`na.action`	A function which indicates what should happen when the data contain `NAs`. The default is set by the `na.action` setting of `options`, and is `na.fail` if that is unset. The "factory-fresh" default is `na.omit`. Another possible value is `NULL`, no action. Value `na.exclude` can be useful.
`allow.dup`	If `TRUE` the algorithm will not stop when duplicated breakpoints are found. The output will contain a slot called `dup`, a data.frame with the breakpoint number of the duplicates (`dup.no`) and the breakpoint number of the its clone among real breakpoints (`pts.no`)
`not.inj.action`	What should be done when `f: y -> x` is not injective.
`sort.data`	Should the data be sorted according to `x` values (when they are not) before fitting the broken-stick model.
`...`	Further arguments to be passed to `bsmfit` such as `eco.mem`.

See bsmfit, the function called by brokenstick on each iteration to fit the model with specified points.

A bsm object with (depending on eco.mem):

pts.x The x values of the brokenstick points (eco.mem inefficient).
pts.y The y values of the brokenstick points (eco.mem inefficient).
slope The slopes of each "stick" of the model (eco.mem inefficient).
intercept The intercept of each "stick" of the model (eco.mem inefficient).
pts (eco.mem < 4) The index of the brokenstick points.
na.action Information from the action which was applied to object if NAs were handled specially. (eco.mem inefficient)
residuals (eco.mem < 3) The model's residuals.
fitted.values (eco.mem < 2) The fitted values.
data (eco.mem < 1) The input data used for fitting.
pts.no The iteration number of points. (eco.mem inefficient)

optBrokenstick and predict.bsm, residuals.bsm, update.bsm, summary.bsm, coef.bsm, plot.bsm, as.data.frame.bsm for other functions with a S3 method for bsm objects.

data(exses)
dv <- tdrply(identity, 1:2, no = 100, obj = exses)[[1]]

# Syntax
bsm <- brokenstick(dv$time, dv$depth)
bsm <- brokenstick(dv) # if two columns
bsm <- brokenstick(depth ~ time, dv) 
bsm <- with(dv, brokenstick(depth ~ time))