In this tutorial, we describe how to include a new emulator to the surmise's framework. We illustrate this with PCGP--an emulator method located in the directory \emulationmethods.
In surmise, all emulator methods inherit from the base class :pysurmise.emulation.emulator. An emulator class calls the user input method, and fits the corresponding emulator. :pysurmise.emulation.emulator.fit and :pysurmise.emulation.emulator.predict are the main :pysurmise.emulation.emulator class methods. It also provides the functionality of updating and manipulating the fitted emulator by :pysurmise.emulation.emulator.supplement, :pysurmise.emulation.emulator.update, and :pysurmise.emulation.emulator.remove class methods.
In order to use the functionality of the base class :pysurmise.emulation.emulator, we categorize the functions to be included into a new emulation method (for example PCGP) into two categories.
:pyfit and :pypredict are the two obligatory functions for an emulation method. :pyfit takes the inputs X, parameters θ, and the function evaluations f, where X ∈ ℝN × p, θ ∈ ℝM × d, and f ∈ ℝN × M. In other words, each column in f should correspond to a row in θ. Each row in f should correspond to a row in X. In addition, the dictionary fitinfo is passed to the :pyfit function to place the fitting information once complete. This dictionary is used keep the information that will be used by :pypredict below.
The :pyPCGP.fit is given below for an illustration:
PCGP
fit
Once the base class :pysurmise.emulation.emulator is initialized, :pysurmise.emulation.emulator.fit method calls the developer's emulator's :pyfit function, and places all information into the dictionary fitinfo.
predict
:pysurmise.emulation.emulator.predict method returns a prediction class object, which has methods :pysurmise.emulation.prediction.mean, :pysurmise.emulation.prediction.mean_gradtheta, :pysurmise.emulation.prediction.var, :pysurmise.emulation.prediction.covx, :pysurmise.emulation.prediction.covxhalf, :pysurmise.emulation.prediction.covxhalf_gradtheta, :pysurmise.emulation.prediction.rnd, and :pysurmise.emulation.prediction.lpdf.
Those expressions are defined within the base class to simplify the usage of the fitted models. In order to use those methods, the emulation method developers should either include functions :pypredictmean, :pypredictmean_gradtheta, :pypredictvar, :pypredictcovx, :pypredictcovxhalf, :pypredictcovxhalf_gradtheta, :pypredictrnd, and :pypredictlpdf into their methods, or define within the dictionary fitinfo using the keys mean, mean_gradtheta, var, covx, covxhalf, covxhalf_gradtheta, rnd, and lpdf.
supplementtheta() is an optional function for an emulation method.