RSA Vectorize samples for convergence analysis

[RyanVries]

Is your feature request related to a problem? Please describe.

Convergence analysis using the mean and SD sensitivities of the dummies requires rerunning the function with a different number of samples, which takes more time.

Describe the solution you’d like

Allow the input of a vector for samples, then run the RSA analysis as normal for the maximum number of samples in the vector. Subsequently, take subsets (sizes equal to the values in the samples vector) of the simulations and corresponding sensitivities to compute the mean and SD of the dummies at each instance. Thereby enabling the assessment of the convergence of the dummy sensitivities without repeatedly having to rerun the code.

Describe alternatives you’ve considered

Additional context