Create self-contained example projects that can be used to learn the
calibrar workflow, test installation, or prototype calibration code.
Each demo writes a small set of files to disk and returns the information
needed to read the setup table, load the observed data, define the model
interface, and run a calibration.
Usage
calibrar_demo(path = NULL, model = NULL, ...)
# S3 method for class 'calibrar.demo'
print(x, ...)Arguments
- path
Directory where the demo project should be created. If
NULL, the current working directory is used.- model
Name of the demo model to generate. Supported values currently include
"PoissonMixedModel","PredatorPrey","SIR", and"IBMLotkaVolterra".- ...
Further arguments passed to
print().- x
A
calibrar.demoobject returned bycalibrar_demo().
Value
An object of class calibrar.demo, which also inherits from
calibrar.results(). The returned list contains the demo project path,
the true parameter values used to generate the synthetic data, the path to
the calibration setup file, suggested starting values and bounds, phase
information, model-specific constants, and a precomputed objective value at
the true parameter set.
Details
The generated project is meant to be a compact, runnable example of the
calibrar workflow rather than a realistic scientific application. The
returned object typically contains enough information to proceed through the
standard helper functions:
calibration_setup()to read the setup table,calibration_data()to load the observed data,calibration_objFn()to build the objective function, andcalibrate()oroptim2()to run the optimisation.
The exact files created depend on model, but they usually include a
calibration settings table, one or more observation files, and any required
forcing or master data files.
The currently available demos are:
"PoissonMixedModel"A Poisson autoregressive mixed model for the dynamics of a population observed at several sites over time.
"PredatorPrey"A deterministic Lotka–Volterra predator–prey system solved numerically.
"SIR"A simple susceptible–infected–recovered epidemic model solved numerically.
"IBMLotkaVolterra"A stochastic individual-based predator–prey example.
Examples
tmp = tempdir()
demo = calibrar_demo(path = tmp, model = "PoissonMixedModel", T = 8, L = 3)
#> Creating observed data list for calibration...
#> Loaded observed data for variables: 'site_1', 'site_2', 'site_3'.
print(demo)
#> Calibration demo using model 'PoissonMixedModel'.
#> Target value: -6988.968
#> Parameters:
#> alpha beta gamma1 gamma2 gamma3 gamma4
#> 0.400000000 -0.400000000 -0.011421355 0.100721594 0.217153872 -0.138190768
#> gamma5 gamma6 gamma7 sd mu_ini1 mu_ini2
#> -0.256919871 0.009345234 -0.047141311 0.200000000 4.330733340 4.624972813
#> mu_ini3
#> 4.204692619
#> Lower bounds:
#> alpha beta gamma1 gamma2 gamma3 gamma4 gamma5 gamma6 gamma7 sd
#> 0.0 -0.5 -1.0 -1.0 -1.0 -1.0 -1.0 -1.0 -1.0 0.0
#> mu_ini1 mu_ini2 mu_ini3
#> 3.7 3.7 3.7
#> Upper bounds:
#> alpha beta gamma1 gamma2 gamma3 gamma4 gamma5 gamma6 gamma7 sd
#> 1.0 0.5 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0
#> mu_ini1 mu_ini2 mu_ini3
#> 5.0 5.0 5.0
#> Phases:
#> alpha beta gamma1 gamma2 gamma3 gamma4 gamma5 gamma6 gamma7 sd
#> 1 1 1 1 1 1 1 1 1 NA
#> mu_ini1 mu_ini2 mu_ini3
#> NA NA NA
#> Guess (starting point):
#> alpha beta gamma1 gamma2 gamma3 gamma4 gamma5 gamma6 gamma7 sd
#> 0.200 0.100 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.200
#> mu_ini1 mu_ini2 mu_ini3
#> 4.431 4.554 4.357
#>
#> Files location: C:\Users\ricar\AppData\Local\Temp\RtmpMT8afP/PoissonDemo
setup = calibration_setup(demo$setup)
observed = calibration_data(setup, path = demo$path)
#> Creating observed data list for calibration...
#> Loaded observed data for variables: 'site_1', 'site_2', 'site_3'.