otusecases
A repository of OpenTURNS test cases
What is it?
This project contains OpenTURNS use cases and datasets.
- The datasets can be used to perform data analysis with OpenTURNS. The goal is to have a set of .csv files which can be easily imported with OpenTURNS for example to fit a marginal distribution, to fit a copula or to create a datamodel.
- Each use-case is made of a function and the distribution of its inputs. Each use-case has a specific methodological goal: central dispersion, reliability, sensitivity analysis or calibration. In general, each use-case is presented in a Notebook: the equations of the function, the references (if any), the simplest possible study which shows how to use the test-case. The use-cases can be used to benchmark a method, but the Notebook only shows the simplest (e.g. the Monte-Carlo method): the goal of each example is not to actuall show the benchmark results. Generally, each use case has a specific goal (e.g. central dispersion), but some use cases can be derived to achieve different methodological goals.
It is based on OpenTURNS.
The scripts are tested with OT 1.13.
Overview
- Axial stressed beam (reliability)
- Cantilever beam (reliability)
- Perrin case (sensitivity analyis)
- Chaboche model (calibration)
- Viscous vertical fall (calibration)
- Flooding case (central dispersion, calibration)
- R-S case (reliability)
- Deflection of a tube (reliability)
- G-Sobol' function (sensitivity analysis)
- Ishigami function (sensitivity analysis)
- Logistic model (calibration, stochastic process)
- Morris function (sensitivity analysis)
- Nonlinear oscillator (reliability)
- Product function (sensitivity analysis)
How to install?
Requirements
The dependencies are:
- Python >= 3.3
- OpenTURNS >= 1.13
- Jupyter Notebook
- otmorris
Installation
Using the latest python version is prefered! Then to install:
git clone https://github.com/mbaudin47/otusecases.git cd otusecases