[ot-users] Sample transformation

Pamphile ROY roy at cerfacs.fr
Sun Dec 3 12:31:26 CET 2017


Hi Sofiane, 

Thanks for the instantiation mitigation. 
It seems to be working as expected for now using both OT1.9 and OT1.10. 

Sincerely, 

Pamphile 


De: "HADDAD Sofiane" <sofiane_haddad at yahoo.fr> 
À: "regis lebrun" <regis_anne.lebrun_dutfoy at yahoo.fr>, "roy" <roy at cerfacs.fr> 
Cc: "users" <users at openturns.org> 
Envoyé: Vendredi 1 Décembre 2017 13:22:47 
Objet: Re: [ot-users] Sample transformation 

Hi all, 

I think there is something to fix. 
This exception is probably due to [ https://github.com/openturns/openturns/commit/aa046883e4d4d2288c2921961fb4a4d48b39c085 | Fix chaos memory usage · openturns/openturns at aa04688 ] 

[ https://github.com/openturns/openturns/commit/aa046883e4d4d2288c2921961fb4a4d48b39c085 | Fix chaos memory usage · openturns/openturns at aa04688Add accessors to the ProjectionStrategy interface to fix the reference counting in FunctionalChaosAlgorithm. Fix... ] 


However I think that when using OT 1.9 with your script, there is also another problem : FunctionalChaosAlgorithm sets a FixedExperiment and override the GaussProductExperiment You can see it if you add pc_algo to the output of your function and get its projection strategy 

BTW, you can change the instanciation of the the class as follows : 


pc_algo = ot.FunctionalChaosAlgorithm(inputSample, weights, column, 
distribution, adaptiveStrategy) 
pc_algo.setProjectionStrategy(projectionStrategy) 


This will sets the projection you defined without any exception 

Bien cordialement, 
Sofiane HADDAD 


Le mercredi 29 novembre 2017 à 18:05:19 UTC+1, roy < roy at cerfacs.fr > a écrit : 


Hi Regis, 

Thanks for your reply. 

I am using the same script from this conversation. 
I enclosed it again. 

Sincerely, 

Pamphile 





Le 29 nov. 2017 à 08:15, regis lebrun < [ mailto:regis_anne.lebrun_dutfoy at yahoo.fr | regis_anne.lebrun_dutfoy at yahoo.fr ] > a écrit : 

Hi Pamphile, 

Could you please provide us a script that reproduce the error? It looks very strange, as this class is tested in more than 10 unit tests and is extensively used in industrial studies. BTW, ot.ComposedDistribution is by no means restricted to the independent copula as your comment could suggest. 

Please note that it is not the physical space distribution which is checked by the GaussProductExperiment class, 
but the distribution defining the functional basis. You should use either OrthogonalProductPolynomialFactory or OrthogonalProductFunctionFactory to build your multivariate basis from 1D orthogonal bases to insure that the resulting multivariate distribution has an independent copula. 

Please give me a feedback on this problem ASAP. 

Cheer 

Régis 


Le mardi 28 novembre 2017 à 23:01:53 UTC+1, roy < [ mailto:roy at cerfacs.fr | roy at cerfacs.fr ] > a écrit : 


Hi Regis, 

On the 1.10, I get this error calling the FunctionalChaosAlgorithm: 

File "/Users/roy/Applications/miniconda3/envs/batman3/lib/python3.6/site-packages/openturns/metamodel.py", line 3849, in __init__ 
this = _metamodel.new_FunctionalChaosAlgorithm(*args) 
TypeError: InvalidArgumentException : Error: the GaussProductExperiment can only be used with distributions having an independent copula. 

But this was working on 1.9. I do not understand the issue as the distribution is an ot.ComposedDistribution. I tried to explicitly add ot.IndependentCopula 
without any change. 

Thanks in advance, 


Pamphile ROY 
Chercheur doctorant en Quantification d’Incertitudes 
CERFACS - Toulouse (31) - France 
[ callto:+33 (0) 5 61 19 31 57 | +33 (0) 5 61 19 31 57 ] 
[ callto:+33 (0) 7 86 43 24 22 | +33 (0) 7 86 43 24 22 ] 





_______________________________________________ 
OpenTURNS users mailing list 
[ mailto:users at openturns.org | users at openturns.org ] 
[ http://openturns.org/mailman/listinfo/users | http://openturns.org/mailman/listinfo/users ] 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://openturns.org/pipermail/users/attachments/20171203/89cb76aa/attachment.html>


More information about the users mailing list