Use rp_measures
A distance matrix (set emRad = NA
to estimate a radius), or a matrix of zeroes and ones
Threshold for distance value that counts as a recurrence (ignored is RM
is a binary matrix)
Minimal diagonal line length (default = 2
)
Minimal vertical line length (default = 2
)
Minimal horizontal line length (default = 2
)
Maximal diagonal line length (default = length of diagonal -1)
Maximal vertical line length (default = length of diagonal -1)
Maximal horizontal line length (default = length of diagonal -1)
Auto-recurrence? (default = FALSE
)
Use a theiler
window around the main diagonal (Line of Identity/Synchronisation) to remove auto-correlations at short time-lags:
0
will include the main diagonal in all RQA measure calculations.
1
will remove the main diagonal from all RQA measure calculations.
NA
(default), will check if the matrix is symmetrical , if so, it will remove the diagonal by setting theiler = 1
(Line of Identity, Auto-RQA), if it is not symmetrical (Line of Synchronisation, Cross-RQA) it will set theiler = 0
.
A value greater than 1
will remove that many diagonals around and including the diagonal from all RQA measure calculations. So theiler = 2
means exclude 2
diagonals around the main diagonal, including the main diagonal itself: [-1,0,1]
.
If theiler
is a numeric vector of length(theiler) == 2
it is possible to exclude an asymmetrical window. The values are interpreted as end points in a sequence of diagonal ID's, e.g. theiler = c(-1,5)
will exclude [-1,0,1,2,3,4,5]
. If length(theiler) > 2
, the values will be considered individual diagonal ID's, e.g. theiler = c(-3,-1,0,2,5)
, will exclude only those specific ID's. Also see the note.
Force chromatic RQA? If NA
the value of the RM
attribute "chromatic"
will be used, if present (default = NA
)
Return matrices? (default = FALSE
)
How many bootstrap replications? (default = NULL
)
Confidence limit for bootstrap results (default = .95
)
Other Recurrence Quantification Analysis:
rp_cl()
,
rp_measures()