lookup_neighbour_trajectories: Find neighbouring trajectories
In rossklin/dynpan: Dynamics estimation for multiple time series

Find neighbouring trajectories of points in a data.table

1
2
3

lookup_neighbour_trajectories(tt, points, k, timesteps = NULL, times = NULL,
  only.indices = TRUE, trajectory.distance.name = "distance",
  time.distance.name = "relative")

`tt`	`time.table` to find trajectories in
`points`	points close to which to find trajectories
`k`	number of nearest neighbours to base trajectories on
`timesteps`	timesteps around the closest neighbours to include in the trajectories
`times`	relative time around the neighbours to include in the trajectories
`only.indices`	whether to only include index and time variabels in the result (rather than also including measurement and auxiliary variables)
`trajectory.distance.name`	name to use for column containing distance to original neighbouring point (defaults to `"distance"`)
`time.distance.name`	name to use for the time difference of a point in a trajectory to the original neighbouring point (always measured in the time unit of `tt`)

Finds trajectories by first finding the k nearest neighbours to each point in points and picks a temporal sequence around each of them.

At least one of timesteps or times has to be specified, if only the former is specified the value of times is computed using the time delta of tt.

rossklin/dynpan documentation built on May 27, 2019, 11:39 p.m.