Joseph Lizier
|
66676a2692
|
A rewrite of parts of the linear Gaussian conditional MI estimator, in order to more deftly handle linear dependencies within variables 1 and 2, and between each of them and the conditional. Previously we removed sub-variables of those arbitrarily, now the removal is targetted to those sub-variables that are problematic. Also required some variable name changes to reflect different ordering of how variables and the conditionals are concatenated in arrays.
|
2019-03-28 23:58:53 +11:00 |
|
jlizier
|
a34bce25ba
|
Enabled auto-embedding for TE Gaussian. Did so by bringing up auto-embedding for TE functionality up from the KSG estimator to the via Conditional MI class. Also added new property to Gaussian TE for number of surrogates to use for bias correction in embedding (if not using analytic bias correction), like Gaussian AIS. This also necessitated adding an empty preFinaliseAddObservations() method to the TE Gaussian multivariate estimator, to ensure it did not yet try to autoembed until that is enabled in the multivariate calculator. Unit tests included here as well (included using more samples for TE KSG autoembedding to ensure we get the correct answer more reliably -- should probably add a stronger effect for that too).
|
2018-10-17 13:49:09 +11:00 |
|
joseph.lizier
|
fb4d6c423c
|
Javadocs made release-ready in the continuous.gaussian package.
|
2014-08-13 02:24:02 +00:00 |
|
joseph.lizier
|
7803966168
|
Adding GPLv3 license statement to all code headers in the infodymamics.measures.continuous.gaussian, kernel and kozachenko packages (continuing ...)
|
2014-08-06 05:37:59 +00:00 |
|
joseph.lizier
|
b50384a302
|
Added methods for computeSignificance analytically (i.e. without bootstrapping) to the new Gaussian TE calculators (i.e. the Granger causality calculators), plus for directly assigning the joint covariance matrix for each calculator, and a debug method to access the computed determinants.
|
2014-04-23 02:07:29 +00:00 |
|
joseph.lizier
|
fd9bf116f9
|
Added TransferEntropyCalculatorViaCondMutualInfo to allow calculation of TE via any implementation of a conditional MI calculator. Added concrete implementations using this for Gaussian estimators (Gaussian TE estimator is equivalent to Granger causality). Replaced existing Kraskov TE calculator (which used two MI calculators) with one which uses one conditional MI estimator via this new method. Also patched TransferEntropyCommon.computeStartAndEndTimePairs
|
2014-04-07 05:31:52 +00:00 |