Hi, 

This is not possible yet, but it's an easy enhancement: 
https://github.com/openturns/openturns/pull/351/commits/f02be8ee706d636bb2a0b1b1edf9568cf7cc2889
 

What is your use-case exactly ? (ie how would you use the experiment variable) 

j 

> De: "Талех Агасиев" <[email protected]>
> À: "users" <[email protected]>
> Envoyé: Jeudi 27 Octobre 2016 21:47:56
> Objet: [ot-users] Python wrapper

> Hi,

> How can I use my own experiment implementation within ot.Experiment class?
> In other words, I'd like to override the ExperimentImplementation methods from
> Python.

> Something like that:

> import numpy as np
> import openturns as ot

> class MyDoeTechnique(ot.ExperimentImplementation):
> def __init__(self):
> super(MyDoeTechnique, self).__init__()

> def generate(self):
> return ot.NumericalSample(np.random.random((5, 2)))

> experiment = ot.Experiment(ot.Axial(2, [1]))
> experiment.generate()

> experiment = ot.Experiment(MyDoeTechnique())
> experiment.generate() # Not yet implemented exception Is there some wrapper
> class to pass extrenal code throw the SWIG ( like OpenTURNSPythonFunction for
> NumericalMathFunction) ?
> Thank you for helping out !

> Best regards,
> Taleh

> _______________________________________________
> OpenTURNS users mailing list
> [email protected]
> http://openturns.org/mailman/listinfo/users

-- 
Julien Schueller 
Phimeca Engineering 
www.phimeca.com 
_______________________________________________
OpenTURNS users mailing list
[email protected]
http://openturns.org/mailman/listinfo/users

Reply via email to