Compute corrected rediscretized sinuosity index for a given tracklet.

Given a data frames containing tracking information for a given tracklet, this function rediscretize the tracklet and returns a vector containing the value of sinuosity along the trajectory.

Usage

sinuosity(df, scale = 1, segL = NULL, timeCol = "frame", compass = NULL)

Arguments

df: A data frame containing x and y coordinates in columns named "x.pos", "y.pos" for a given tracklet, as well as a column containing time information, whatever the unit, over the tracklet.
scale: A ratio corresponding to the scaling factor which should be applied to the trajectory coordinates. (e.g., size in cm / size in pixels; see TrajScale, default = 1.
segL: A numeric value expressed in the unit specified by user and corresponding to the length of the resampling step needed to discretize the input trajectory (optional) see TrajRediscretize.
timeCol: A character string corresponding to the name of the column containing Time information (default = 'frame').
compass: A value used to specify the compass direction (in radians). If not NULL, turning angles are calculated for a directed walk, otherwise, a random walk is assumed (default = NULL).

Value

This function returns a value of sinuosity for a given tracklet according to TrajSinuosity2 from the trajr package.

References

Benhamou, S. (2004). How to reliably estimate the tortuosity of an animal's path. Journal of Theoretical Biology, 229(2), 209-220. doi:10.1016/j.jtbi.2004.03.016.

Author

Quentin PETITJEAN

Examples


set.seed(2023)
# generate a dummy tracklet
## start to specify some parameters to generate the tracklet
TrackL <-
  100 # the length of the tracklet or a sequence to randomly sample tracklet's length

TrackDatTemp <-
  trajr::TrajGenerate(sample(TrackL, 1), random = TRUE, fps = 1)
TrackDat <-
  data.frame(
    x.pos = TrackDatTemp[["x"]] - min(TrackDatTemp[["x"]]),
    y.pos = TrackDatTemp[["y"]] - min(TrackDatTemp[["y"]]),
    frame = TrackDatTemp[["time"]]
  )

# compute the sinuosity of the particle' trajectory
MoveR::sinuosity(TrackDat, scale = 1, timeCol = "frame")
#> [1] 0.3358906