Bs/stability fixes

[bspanoghe]

P5

Advanced notebook

• Defined prior inside Turing model to prevent world-age issues
• Bounded Pt to be positive to prevent errors when it's used as the mean of a Poisson distribution

P6

All

• Changed solver to AutoTsit5(Rosenbrock23())

Calibration intro

• Slightly add onto intro explanation for MLE vs MAP and MLE/MAP vs MCMC
• Updated parameter explanation: we don't directly specify initial values for optimization
• Add part explaining priors
• Move section "The measurement data" to before prior explanations to better fit flow of the explanation
• Add truncation to priors if necessary
• Change InverseGamma prior to better known Exponential distribution
• Remove observed output variable W_meas from model input and condition it instead (also matching P8)
• Remove instantiations of unused ODEProblems outside of Turing models
• Add chain plots to check convergence
• Remove use of summarize function to extract mean values from chain

Fermenter calibration:

• Remove observed output variable from model input and condition it instead (also matching P8)
• Remove instantiations of unused ODEProblems outside of Turing models
• Changed InverseGamma priors to Exponentials

Irrigation

• Remove observed output variable from model input and condition it instead (also matching P8)
• Changed InverseGamma priors to Exponentials

P8

Model selection intro

• Copied model changes from calibration intro
• Changed solution of exercise 2 to MAP estimation, as requested in the exercise
• Remove instantiations of unused ODEProblems outside of Turing models

Friction notebook

• Changed to match other updates
  • Changed InverseGamma priors to Exponentials
  • Changed solver to AutoTsit5(Rosenbrock23())

[github-actions]

Once the build has completed, you can preview your PR at this URL: https://Kermit-UGent.github.io/ModSim/previews/PR69/

[warclaer]

Good additions

[bspanoghe]

Approved by @gvhaelew 🎊