Régis LEBRUN

I made a small benchmark: ``` import openturns as ot U = ot.Uniform(-14.0, 14.0) f1 = ot.SymbolicFunction("x", "sin(x)") f2 = ot.SymbolicFunction("x", "abs(x)") d2 = ot.CompositeDistribution(f2, ot.CompositeDistribution(f1, U)) d3 = ot.CompositeDistribution(ot.ComposedFunction(f2,...

Here are the result on my laptop using current master: ################################################## d2 pdf= 1077999.7332182592 evals/s d2 cdf= 1404892.5840847993 evals/s d2 rng= 5230548.980777816 rngs/s ################################################## d3 pdf= 529994.7850978897 evals/s d3 cdf=...

Ok but why do we need to actually store the model? The setModel() accessor could simply emit a warning, and the getModel() accessor could build a DatabaseFunction on demand, no?

@jschueller The choice of TNC as the default optimization algorithm is more and more questioning, as we get better results using Cobyla in many places. IMO we should switch to...

If I remember well, this algorithm was presented 10 years ago during a workshop at IHP. I am pretty sure that this article would be of interest https://www.researchgate.net/publication/345206244_Iterative_algorithms_for_non-conditional_and_conditional_simulation_of_Gaussian_random_vectors but unfortunately...

To be more precise: * Do we use 'in the :class:`~openturns.ResourceMap`', 'of the :class:`~openturns.ResourceMap`', 'in :class:`~openturns.ResourceMap`'? * Do we use `*Key-Name*`, `'Key-Name'` or '``'Key-Name'``'? You suggest to use 'entry' for...

Hi @mbaudin47 , There is currently no generic parameter management in distributions and distribution factories, except for the generic ones: MethodOfMomentsFactory, MaximumLikelihoodFactory and LeastSquaresDistributionFactory. You should use one of them...

Right, I have been fouled by the includes in LogNormalFactory.cxx: there is a ``` #include "openturns/LeastSquaresDistributionFactory.hxx" ``` but it is unused, and the least-squares problem solved by a (much more)...

I take this one for myself

@jschueller I already saw it. It implements A.B where A is sparse and B is dense, it should be easy to adapt it to the symmetrical case. My main concern...