Fixed doc

[mbaudin47]

• Fixed enumerate list in doc/examples/sensitivity_methods/plot_benchmark_sensitivity_methods.py.

• Added an example into the README, so that the user can known what the package is about.
• Created an example to show how to create a new reliability benchmark problem.
• Fixed LaTeX equations of sensitivity problems.
• Used Black on Python scripts.
• Updated formula for OpenTURNS 1.26, for reliability problems using ExprTk.

[mbaudin47]

@mbalesdent. In eeef398, I created a new example to show how to create a new reliability benchmark problem. This was a feature you requested a long time ago. The problem is adapted from #69, i.e. the two-degree-of-freedom damped oscillator. The problem is, however, difficult to solve using plain Monte-Carlo. May you suggest another algorithm which better solves the problem?

[mbaudin47]

@jschueller : The Github Actions prints:

writing output... [ 60%] auto_examples/sensitivity_problems/plot_gaussian_product_sensitivity .. user_manual/_generated/otbenchmark.GaussianProductSensitivity

(Error in parallel process)
Traceback (most recent call last):

[...]

UnicodeDecodeError: 'ascii' codec can't decode byte 0xe2 in position 1753: ordinal not in range(128)

This seems to be created by a "ë" in some file, most likely by "Michaël". But I cannot find where. Is there any trick to find the exact location of any Sphinx error?

[jschueller]

indeed you have a non-ascii long minus char (−) inside the omega_p - omega_s formula:

$ LC_ALL=C  grep -PInr "[\x80-\xFF]"  otbenchmark/|grep math
otbenchmark/_BoreholeSensitivity.py:29:        * :math:`T_u`: transmissivity of upper aquifer (m²/yr)
otbenchmark/_BoreholeSensitivity.py:31:        * :math:`T_l`: transmissivity of lower aquifer (m²/yr)
otbenchmark/_NLOscillatorSensitivity.py:65:        * :math:`r = (\omega_p − \omega_s) / \omega_a` a tuning parameter.

the m² are ok since they are outside math

you can also see them in colors with

LC_ALL=C  grep --color='auto' -PI -nr "[\x80-\xFF]"  otbenchmark/

[mbaudin47]

Thank you for pointing this. I can now see it indeed using VSCode:

[mbaudin47]

@m-balesdent Does this correspond to your requirement?

[mbaudin47]

@regislebrun : Can you review this PR and tell if it corresponds to what you expect?

[mbaudin47]

@jschueller : I suggest to review this PR and merge it, if you are ok with that.

[mbaudin47]

@jschueller : Thank you very much and happy new year!