Log-uniform sampling for global sensitivity analyses

Hello,

When using the sbiosobol function, what's the default distribution that is used to sample between the bounds? Is there a way to specify uniform distributions for some parameters and log-uniform for others? I know that this can be done by changing the structural model so that the log-uniform parameters are used e.g. as 10^k, but I am wondering if there's a way to do this without modifying the model?

Thank you,

Abed

Answer 1

[Jeremy Huard]

Hi Abed,

This option was added to sbiosobol in the latest release R2021b.

Best regards,

Jérémy

[Abed Alnaif]

Hi Jeremy,

Thanks for letting me know. I'll update my version, then. Can you please let me know how to use this option (I can't find it in the R2021b help for sbiosobol)?

[Jeremy Huard]

Hi Abed,

You can specify such samples by passing in a SimBiology.Scenarios object as the second input argument or by specifying the 'Distributions' name-value argument with a vector of probability distribution objects as its value.

https://www.mathworks.com/help/releases/R2021b/simbio/ref/sbiosobol.html#mw_1ef1927b-eefd-499b-a88f-8c3542d4df85

[Abed Alnaif]

Thank you, Jeremy!