Michael Baudin

Analysis of residuals : the blue curve should be renamed. - LeastSquares : the blue label should be named "Normal, after calibration" ; - Bayesian : the blue label should...

Below are some useful links : - [~60x speed-up of Linux "perf"](https://eighty-twenty.org/2021/09/09/perf-addr2line-speed-improvement) - [How do I compile a Linux perf tool with all features? For Linux 4.0 on Ubuntu.](https://www.quora.com/How-do-I-compile-a-Linux-perf-tool-with-all-features-For-Linux-4-0-on-Ubuntu) -...

Changed 3 years ago by schueller I have a script that does that: It needs enhancements in MaximumLikelihoodFactory (​https://github.com/openturns/openturns/pull/375). ``` class TruncatedFactory(object): def __init__(self, distribution, lb=None, ub=None): self.distribution = distribution...

In order to improve the performances of the Kolmogorov test, I implemented several new build methods in: https://github.com/mbaudin47/openturns/commits/FixChaos with 3 build methods: * buildMethodOfScaledLikelihoodMaximization is the current method based on...

We may find an appropriate reference to the bibliography.

The following is a draft. [casCrue-4vars-SRC-with-precision.txt](https://github.com/openturns/openturns/files/3007638/casCrue-4vars-SRC-with-precision.txt) This uses the flooding test case and uses the CorrelationAnalysis.SRC function to estimate the indices. These are based on a plain Monte-Carlo sampling with...

@sofianehaddad : you are right - if, during the execution of the incremental algorithm, the data reveals that the model is not linear, the basic hypothesis is not satisfied. What...

Oups, I missed your message: which paper are you interested in?

I have already managed to get some intermediate results, and I am currently working again on this topic. I will let you know when I have some results (the maths...

This is still true. In my example, I consider the distribution which is computed by conditioning the random vector (X0, X1, X2) upon X1=x1. Therefore, the description should be ("X0...