jidt
jidt copied to clipboard
jlizier
JIDT: Java Information Dynamics Toolkit for studying information-theoretic measures of computation in complex systems
Estimator for TE in continuous-time on spike trains or other event-based data. Also includes demo python scripts for evaluating the estimator as well as performing effective network inference.
The code uses 2/k but I think this should be 1/k, in line with the correction of (m-1)/k for m continuous variables referred to in the MI estimator in [this...
Had this reported on Mac; have not checked other. Would likely be an issue with the file save call (but I'm surprised it happens at all - maybe requires escaping...
It seems it needs the 1D/2D specific methods to work.
Greetings! I am using Python. I followed the tutorial using jpype. The code below works when I set the histories and delays to 1. My code: ` cteClass = jpype.JPackage("infodynamics.measures.continuous.kraskov").ConditionalTransferEntropyCalculatorKraskov...
Dear Prof. Lizier, Code and wiki page pull request to finalise capstone submission. I'm not sure whether the links in the wiki page will continue to work after you've accepted...
Base is confusing, it conveys that it's the base of the logarithm. Something about alphabet size will be more direct.
Based on a question raised during CNS2020 conference: To what extend is it possible to use the tool-set in a streaming manner? We are currently building a multi-scale simulator framework...
Extension to MI calculator KSG Alg 1 for handling data with periodic boundary conditions. Using underlying data structure, v-p tree.
I am not sure if this is not by design for some reason, but if I initialise e.g. ActiveInfoStorageCalculatorKraskov with an embedding length of 3, calculator.initialise(3, 1), and add observations...
