integrate: Integrator of a marked point process
In ppstat: Point Process Statistics

Description Usage Arguments Details Value Author(s) See Also Examples

Computes the integral with respect to a marked point process. Either the integrated process with jump sizes as weights is computed, or the integrated process with respect to the counting process is computed.

1 2	## S4 method for signature 'MarkedPointProcess' integrator(object, f = 1, jumpVar = '', result = 'JumpProcess', ...)

`object`	a `MarkedPointProcess` object.
`f`	a `function` or `numeric` vector. The integrand, which is integrated with respect to the marked point process. If `f` is a function it needs to be vectorized in its first argument. The default value is `1`, meaning a constant integrand.
`jumpVar`	a `character`. If not equal to one of the column names of `getMarkValue(object)` the counting process for `mark` is computed. The default value is `''`.
`result`	a `character`. By default, the function returns an object of class `JumpProcess`. If `result = 'numeric'` it returns instead the integrated process as a numeric vector.
`...`	other arguments passed to `f`.

The integral computed is with respect to all points in the marked point process. If the integral with respect to the process having only a subset of the marks is needed, use the subset function on the object to select the appropriate subset of marks.

It is possible to integrate functions that depend upon marks or mark values. This can be done either by precomputing the function evaluations, or by specifying f to take more than one argument, so that it works correctly as a vectorized function, and supply the additional arguments in the call of integrator.

A JumpProcess object or a numeric vector. Note that some information is lost when returning just a numeric, e.g. the unit identifications.

Niels Richard Hansen, Niels.R.Hansen@math.ku.dk.

jumpProcess, MarkedPointProcess, JumpProcess, subset.

MP <- data.frame(time = seq(5, 95, 10),
                 markType = rep("foo", 10),
                 jump = rnorm(10))

## Not run: ## The counting process only
plot(integrator(markedPointProcess(MP, 1:100)))

## The integrated pure jump process.
plot(integrator(markedPointProcess(MP, 1:100), jumpVar = 'jump'))

## With an integrand
plot(integrator(markedPointProcess(MP, 1:100), f = cos))

## End(Not run)