Michael Baudin

Results 190 comments of Michael Baudin

Soit $F_{n, \alpha}$ la loi binomiale de paramètres $n$ et $\alpha$. ![image](https://github.com/openturns/openturns/assets/31351465/70ec45ac-fe5a-4921-88bd-6b56ad844a90) ![image](https://github.com/openturns/openturns/assets/31351465/60b1c3be-ca65-4207-a2be-8742558ad417) **Figure 1.** Fonction de répartition de la loi binomiale de paramètres n et α. ![image](https://github.com/openturns/openturns/assets/31351465/7cfb9e2c-1ff5-47ba-92ea-ff974714619d) **Figure 2.**...

For any discrete random variable with cumulative distribution function $F$, let $k = F^{-1}(\beta)$ be the quantile of level $\beta$. Therefore: $$ F(k - 1) < \beta \leq F(k). $$...

Scary bug! For my own information, I tuned the algorithm so as to get the correct order of magnitude: ```python import openturns as ot from openturns.usecases import cantilever_beam cb =...

To make the review easier, I write below a list of the examples which are changed by the PR: - [plot_estimate_non_parametric_distribution.html](https://openturns.github.io/openturns/latest/auto_data_analysis/distribution_fitting/plot_estimate_non_parametric_distribution.html) - [plot_overfitting_model_selection.html](https://openturns.github.io/openturns/latest/auto_meta_modeling/general_purpose_metamodels/plot_overfitting_model_selection.html) - [plot_optimization_pagmo.html](https://openturns.github.io/openturns/latest/auto_numerical_methods/optimization/plot_optimization_pagmo.html) - [plot_field_fca_sobol.html](https://openturns.github.io/openturns/latest/auto_reliability_sensitivity/reliability_processes/plot_field_fca_sobol.html)

Here is the PDF: ![output](https://github.com/openturns/openturns/assets/31351465/a7af9271-346d-4a0d-9cd5-72fea4b57f97) It is true, indeed, that the value x = 1.5 is so that f(0) = 0. But why should this be related to the definition...

You are right. This is done at: https://github.com/openturns/openturns/blob/4fc2024d0cbccb92b4c5f70df8fd36598c903eb5/lib/src/Uncertainty/Model/DistributionImplementation.cxx#L2307 Actually, we do not have the required information in the `Distribution`: is the CDF continuous and strictly increasing? If yes, then `Brent`...

- That would be a good reference from importance_form. - A new help page would be dedicated to the importance factors of a linear model. Then the FORM and Taylor...

> I just wanted to say that I do not agree to mention da Veiga about the importance factors as da Veiga is not the author of these indices and...

@josephmure : I took your suggestions into account. Please review the changes.

@josephmure : Please review this PR.