RQA

[sultornsanee]

After I create recurrence matrix using "crp" function and I get a recurrence matrix, can I calculate RQA from this matrix?

I read your manual but it seems all function needed "time series" as input.

[Norbert]

yes, you can do it. for example:

Code: Select all

% create recurrence plot
X = crp(timeseries,'silent');

% get distributions of line lengths
% (actually you should apply here the Theiler window before, but I omit it for simplification)
[m l] = dl(X); % l is a list of all line lengths of diagonal lines
[m v] = tt(X); % v is a list of all line lengths of vertical lines

% recurrence rate
mean(X(:))

% determinism
sum(l(l>=lmin))/sum(X(:)) % where lmin is the minimal diagonal line length, e.g. lmin=2

% mean diagonal line length
mean(l(l>=lmin))

% maximal diagonal line length
max(l(l>=lmin))

% entropy
histL=hist(l,1:length(X));
entropy(histL(:));

for the vertical line based measures (laminarity, trapping time etc.) just use vector v instead of l

best
norbert

[sultornsanee]

Thank you very much ... I will try to test your codes.

[Parri-Dee]

After I create recurrence matrix using "crp" function and I get a recurrence matrix, can I calculate RQA from this matrix?