Example Usage

Contents:

• Fitting a polynomial with Gaussian priors
  • Define synthetic truth and use it to create noisy observations
  • Assume diagonal observation error covariance matrix and define perturbed observations
  • Define Gaussian priors
  • Run forward model in parallel
  • Pick responses where we have observations
  • Condition on observations to calculate posterior using both ES and SIES
  • Plots to compare results
• Estimating parameters of an anharmonic oscillator
  • Setup
  • Plot observations
  • Create and plot prior
  • A single update step
  • Iterative smoother
  • ES-MDA (Ensemble Smoother - Multiple Data Assimilation)
• Linear regression with ESMDA
  • Import packages
  • Create problem data
  • Solve the maximum likelihood problem
  • Solve using ESMDA
  • Plot and compare solutions
• Adaptive Localization
  • Import packages
  • Create problem data
  • Create problem data - sparse tridiagonal matrix \(A\)
  • Solve the maximum likelihood problem
  • Solve using ESMDA
  • Plot and compare solutions
  • Solve using AdaptiveESMDA
  • Correlations between true parameters and solution means
  • Run on several ensemble sizes and seeds

• Index