I = 0.5 * log(RSS_restricted / RSS_unrestricted), which equals
transfer entropy exactly for Gaussian processes. Mirrors
morie.fn.granci and the Gaussian arm of morie.fn.trnfen.
Usage
morie_transfer_entropy_gaussian(x, y, lag = 1L)
Arguments
- x
Source series.
- y
Target series.
- lag
History length.
Value
List with mi, statistic, p_value, df, n.
References
Barnett L, Barrett AB, Seth AK (2009). Physical Review
Letters 103(23), 238701.
Examples
x <- sin(1:60)
y <- c(0, 0.8 * x[-60]) + 0.3 * cos(3 * (1:60))
morie_transfer_entropy_gaussian(x, y)
#> $mi
#> [1] 1.118756
#>
#> $statistic
#> [1] 132.0132
#>
#> $p_value
#> [1] 1.486313e-30
#>
#> $df
#> [1] 1
#>
#> $n
#> [1] 60
#>
