Error No Sobol Indices Available to Plot GlobalSensitivityAnalysisApp Simbiology
Hello,
I recently downloaded the GlobalSensitivityAnalysisApp. I am trying to perform a GSA using a simbiology model. I try running the App, but I get the error "No Sobol Indices Available to Plot. Configure the Sobol section and click the Compute button." I'm not sure if there is a step I'm overlooking. I'm following the instructional video for this app on the Matlab website, and I'm not sure what Configure the Sobol section is referring to. Any guidance would be greatly appreciated.
9 Comments
Time Descending
Did you define the output for the GSA? I can imagine that if the output is not defined, it won't calculate any indices.
Did you define the output for the GSA? I can imagine that if the output is not defined, it won't calculate any indices.
Hi Sietse,
Thank you for replying so quickly. I defined the model and the dose, and in the GSA App, I defined the Observables under Variance decomposition. Is this what you are referring to?
Hi Brett, I'm a little in the dark as to how you've got it set up but can you make sure you've done the following:
- start the app by calling on the MATLAB command line:
startGlobalSensitivityAnalysisApp(<model object name>,<dose object name>)
If you don't add a dose and your initial conditions are all zero, the model will not be perturbed and your Sobol results will likely not mean much.
- In the app, make sure you've selected one or more parameters (I selected k1,k2,k3 in the screenshot I attached)
- In the app, make sure you select an suitable output, either an observable, species or non-constant parameter.
- Make sure you click "Compute" before you click "Plot"
Let me know if after trying this, you still have the same error.
Best,
Sietse
Hi Sietse,
Yes, I set up GSA as you described. I found the names of my doses using "m1.Doses". I specified the dose using "dose = sbiodose('dose_DOX0.565')". My model was named "m1", so I ran "startGlobalSensitivityAnalysisApp(m1, dose)."
I chose one parameter (EC50_DOX), and one observation (Effect.pRB_percent_molecule). I've attached a screenshot below. I hit compute, but I get the same error.
Hi Brett,
there seems to be an issue with your call to startGlobalSensitivityAnalysisApp:
dose = sbiodose('dose_DOX0.565')
This syntax means you are creating a new dose with this name but it does not contain any dosing/target information yet. What you actually want is to select an existing dose in your model. The command for this should read:
dose = getdose(m1,'dose_DOX0.565') % assuming dose_DOX0.565 is the name of your dose
Does it solve the issue?
Best, Jérémy
Hi Jérémy,
Thank you for the suggestion. Yes, this allows me to run the GSA. Thanks!
Strange - we may have to set up a call to go through this. I am afraid I can't see what is going on here.
I see. Do you have an example model and GSA code you could share?
