josephmure
josephmure
Can be used to implement any proposal mechanism in the Metropolis-Hastings algorithm.
Hello, among the various [FunctionalChaosAlgorithm](http://openturns.github.io/openturns/master/user_manual/response_surface/_generated/openturns.FunctionalChaosAlgorithm.html) constructors, we have: * FunctionalChaosAlgorithm(inputSample, outputSample) * FunctionalChaosAlgorithm(inputSample, outputSample, distribution, adaptiveStrategy) I think it would be helpful to have a new intermediate constructor: * FunctionalChaosAlgorithm(inputSample,...
I am not sure what [this line](https://github.com/josephmure/openturns/blob/0ae3e83643240ddac3d7e3fe9dbe0ff9485c8103/lib/src/Uncertainty/Algorithm/MetaModel/Kriging/GeneralLinearModelAlgorithm.cxx#L748) is supposed to do. The block of code immediately above is: ``` const Scalar epsilon = rho_.normSquare(); LOGDEBUG(OSS(false)
The test script [t_Distribution_scipy.py](https://github.com/openturns/openturns/blob/master/python/test/t_Distribution_scipy.py) currently fails Continuous Integration for 2 reasons: 1. The following line is now invalid: https://github.com/openturns/openturns/blob/82c89cd7e5e92aade1b6d216167453508aaa4684/python/test/t_Distribution_scipy.py#L8 It should be replaced with a call to numpy.random.seed(42) 2. SciPyDistributions...
There are several issues related to the way Gibbs sampling works in the [GaussianProcess](https://openturns.github.io/openturns/1.19/user_manual/_generated/openturns.GaussianProcess.html) class. - [x] Using `setSamplingMethod(ot.GaussianProcess.GIBBS)` results in a wrong result. Here is an example where we...
### Bug fix Fixes a bug that could cause an error in the detection of samplers which do not need to recompute the log-posterior density in the case where the...
The docstring of the method is simply "TODO": https://github.com/openturns/openturns/blob/6928d3c65d9efac3784c3f0939df3336e8a896a7/python/src/CovarianceModelImplementation_doc.i.in#L210
A few examples discovered with @JPelamatti : ----------------- ``` In [49]: ot.Interval() == ot.ExpectationSimulationAlgorithm() Out[49]: True ``` Possible explanation: Interval::operator == tries to use the DomainImplementation version https://github.com/openturns/openturns/blob/d6a14afe4954d6f35e7fe79481d918928a45935a/lib/src/Base/Geom/openturns/Interval.hxx#L111 but it...
### ProductFunction constructor One of the `ProductFunction` constructors requires `FunctionImplementation` objects, so the user is required to call the`getImplementation` methods of their `Function` objects. However, after PR #1936 is merged,...
Only a POC: not all RandomVector classes are updated as they should be and the API changes are only done for exploratory purposes. The basic idea is that if we...