info_flow: Information Flow

In rinform: An R Wrapper of the 'Inform' C Library for Information Analysis of Complex Systems

Description

Compute the information flow from lsrc time series src to another ldst time series dst. Optionally, the user can specify lback background time series back to conditioning probabilities.

Usage

info_flow(src, dst, back = NULL, lsrc, ldst, lback = 0)

Arguments

src	Vector or matrix specifying one or more source time series.
dst	Vector or matrix specifying one or more destination time series.
back	Vector or matrix specifying one or more background time series.
lsrc	Integer giving the number of sources.
ldst	Integer giving the number of destinations.
lback	Integer giving the number of backgrounds.

Value

Numeric giving the value of information flow.

Examples

# Information flow across four time series:
ws <- c(0, 0, 1, 0, 1, 1, 0, 1)
xs <- c(0, 0, 1, 0, 1, 1, 0, 1)
ys <- c(0, 0, 1, 0, 1, 1, 0, 1)
zs <- c(0, 0, 0, 0, 0, 0, 0, 0)

info_flow(src = xs, dst = ys, lsrc = 1, ldst = 1)                            # flow ~ 1.0
info_flow(src = xs, dst = ys, back = ws, lsrc = 1, ldst = 1, lback = 1)      # flow ~ 0.0
info_flow(src = ws, dst = zs, back = ys, lsrc = 1, ldst = 1, lback = 1)      # flow ~ 0.0

# Now let's intervene on ys and zs
ws <- c(0, 0, 1, 0, 1, 1, 0, 1)
xs <- c(0, 0, 1, 0, 1, 1, 0, 1)
ys <- c(1, 0, 1, 0, 0, 1, 1, 0)
zs <- c(1, 0, 0, 0, 1, 0, 1, 1)
info_flow(src = xs, dst = ys, lsrc = 1, ldst = 1)                            # flow ~ 0.0
info_flow(src = xs, dst = ys, back = ws, lsrc = 1, ldst = 1, lback = 1)      # flow ~ 0.0
info_flow(src = ws, dst = zs, back = ys, lsrc = 1, ldst = 1, lback = 1)      # flow ~ 1.0

rinform documentation built on April 1, 2018, 12:12 p.m.