plot_snake: Make a 'snake' plot of a Gaussian Process model

In svdataman/gin: Gaussian Processes for time series inference

Description

plot_snake plots the mean and N-sigma band for a GP.

Usage

plot_snake(dat, sigma = c(1), add = FALSE, col.line = 1,
  col.fill = NULL, col.border = NULL, ...)

Arguments

dat	(data frame/matrix) data for time series.
sigma	(float) Plot the +/-sigma band (>0).
add	(logical) if TRUE then start a new plot
col.line	(colour) Colour of the mean line.
col.fill	(colour) Colour of the +/-sigma band.
col.border	(colour) Colour of +/-sigma limits.
...	(anything) any other graphical keywords to be passed to plot(...)

Details

Produces a 'snake' plot. This is a plot showing the mean of a Gaussian process against time (as a thick line) and the variance of the process illustrated as a band. The lines goes through mu(t) and the band covers mu(t) +/- sigma*dy(t) where dy(t) is the standard deviation - this is actually sqrt(diag(covariance_matrix)). Sigma is a constant (>0) to allow one to plot e.g. the +/-2 std.dev band.

Value

None.

See Also

plot_ts

svdataman/gin documentation built on March 12, 2021, 7:37 a.m.