bootpan: Bootstrapping for panel data

Generate bootstrap samples given a sampling procedure for initial values and a procedure for sampling transitions (conditional only on the current value of the process).

1	sample_markov(times, rinitial, rtransition, time.column.name = "time", ...)

`times`	vector of timepoints at which to sample
`rinitial`	function to sample initial data.table/values
`rtransition`	function to sample transitions given a data.frame containing current values and the current and proceeding time point
`time.column.name`	column name to use store time values in (defaults to "time")
`...`	additional values to pass to rinitial and rtransition

rinitial should take no paramters not specified in ... while rtransition takes three arguments: the time to transition from, the time to transition to, and a data.table containing the current set of values, in that order.

rtransition *is* allowed to destructively update the data.table it is passed, but should even then return that destructively updated data table.

rossklin/bootpan documentation built on May 27, 2019, 11:38 p.m.