predictive_info: Predictive Information

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

Description

Compute the predictive information from a time series with history length kpast and future length kfuture.

Usage

predictive_info(series, kpast, kfuture, local = FALSE)

Arguments

series	Vector or matrix specifying one or more time series.
kpast	Integer giving the history length.
kfuture	Integer giving the future length.
local	Boolean specifying whether to compute the local predictive information.

Value

Numeric giving the average predictive information or a vector giving the local predictive information.

Examples

# One initial condition
xs <- c(0, 0, 1, 1, 0, 0, 1, 1, 0)
predictive_info(xs, kpast = 1, kfuture = 2) # 0.985228

# [0.807, 0.807, 1.222, 1.222, 0.807, 0.807, 1.222]
predictive_info(xs, kpast = 1, kfuture = 2, local = TRUE)

# Multiple initial conditions
xs      <- matrix(0, nrow = 9, ncol = 2)
xs[, 1] <- c(0, 0, 1, 1, 0, 0, 1, 1, 0)
xs[, 2] <- c(0, 1, 0, 1, 0, 1, 0, 1, 0)
predictive_info(xs, kpast = 1, kfuture = 2) # 0.244905

# [, 1] -0.515 0.807 -0.363 1.222 -0.515 0.807 -0.363
# [, 2] 0.222 0.485  0.222 0.485  0.222 0.485  0.222
predictive_info(xs, kpast = 1, kfuture = 2, local = TRUE)

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