tests/testthat/test-calibration-pmodel.R

In rsofun: The P-Model and BiomeE Modelling Framework

context("test P-model calibration framework and its parameters")
set.seed(10)

test_that("test GPP calibration routine p-model (BT, likelihood maximization)", {
  skip_on_cran()
  drivers <- rsofun::p_model_drivers
  obs <- rsofun::p_model_validation
  params_fix <- list(
    # kphio              = 0.04998, # setup ORG in Stocker et al. 2020 GMD
    kphio_par_a        = 0.01,  # set to zero to disable temperature-dependence of kphio, setup ORG in Stocker et al. 2020 GMD
    kphio_par_b        = 1.0,
    soilm_thetastar    = 0.6 * 240,  # to recover old setup with soil moisture stress
    soilm_betao        = 0.01,
    beta_unitcostratio = 146.0,
    rd_to_vcmax        = 0.014, # value from Atkin et al. 2015 for C3 herbaceous
    tau_acclim         = 30.0,
    kc_jmax            = 0.41
  )
  
  settings <- list(
    method              = "bayesiantools",
    targets             = c("gpp"),
    sitenames           = "FR-Pue",
    metric              = rsofun::cost_likelihood_pmodel,
    control = list(
      sampler = "DEzs",
      settings = list(
        nrChains = 1,
        burnin = 1,
        iterations = 4
      )
    ),
    par = list(
      kphio = list(lower=0.04, upper=0.09, init=0.05),
      err_gpp = list(lower = 0.01, upper = 4, init = 2)
    )
  )
  
  pars <- rsofun::calib_sofun(
    drivers = drivers,
    obs = obs,
    settings = settings,
    # extra arguments for the cost function
    par_fixed = params_fix,
    targets = c('gpp'),
    parallel = TRUE,
    ncores = 2
  )
  
  # test for correctly returned values
  expect_type(pars, "list")
})

test_that("test GPP calibration routine p-model (GenSA, rmse, all params)", {
  skip_on_cran()
  drivers <- rsofun::p_model_drivers
  obs <- rsofun::p_model_validation
  
  settings <- list(
    method              = "gensa",
    targets             = c("gpp"),
    sitenames           = "FR-Pue",
    metric              = rsofun::cost_rmse_pmodel,
    control = list(
      maxit = 2
    ),
    par = list(
      kphio = list(lower=0.04, upper=0.09, init=0.05),
      kphio_par_a = list(lower = 0, upper = 1, init = 0.2),
      kphio_par_b = list(lower = 10, upper = 40, init =25),
      soilm_thetastar = list(lower = 0, upper = 3000, init = 0.6*240),
      soilm_betao = list(lower = 0, upper = 1, init = 0.2),
      beta_unitcostratio = list(lower = 50, upper = 200, init = 146),
      rd_to_vcmax = list(lower = 0.01, upper = 0.1, init = 0.014),
      tau_acclim = list(lower = 7, upper = 60, init = 30),
      kc_jmax = list(lower = 0.2, upper = 0.8, init = 0.41)
    )
  )
  
  pars <- rsofun::calib_sofun(
    drivers = drivers,
    obs = obs,
    settings = settings,
    optim_out = FALSE,
    # extra arguments for the cost function
    targets = 'gpp'
  )
  
  # test for correctly returned values
  expect_type(pars, "list")
})

test_that("test Vcmax25 calibration routine p-model (BT, likelihood, all params)", {
  skip_on_cran()
  drivers <- p_model_drivers_vcmax25
  obs <- rsofun::p_model_validation_vcmax25
  
  settings <- list(
    method              = "bayesiantools",
    metric              = rsofun::cost_likelihood_pmodel,
    control = list(
      sampler = "DEzs",
      settings = list(
        nrChains = 1,
        burnin = 1,
        iterations = 4
      )
    ),
    par = list(
      kphio = list(lower=0.04, upper=0.09, init=0.05),
      kphio_par_a = list(lower = 0, upper = 1, init = 0.2),
      kphio_par_b = list(lower = 10, upper = 40, init =25),
      soilm_thetastar = list(lower = 0, upper = 3000, init = 0.6*240),
      soilm_betao = list(lower = 0, upper = 1, init = 0.2),
      beta_unitcostratio = list(lower = 50, upper = 200, init = 146),
      rd_to_vcmax = list(lower = 0.01, upper = 0.1, init = 0.014),
      tau_acclim = list(lower = 7, upper = 60, init = 30),
      kc_jmax = list(lower = 0.2, upper = 0.8, init = 0.41),
      err_vcmax25 = list(lower = 0.001, upper = 0.2, init = 0.1)
    )
  )
  
  pars <- rsofun::calib_sofun(
    drivers = drivers,
    obs = obs,
    settings = settings,
    optim_out = FALSE,
    # arguments for cost function
    targets = 'vcmax25'
  )
  
  # test for correctly returned values
  expect_type(pars, "list")
})

test_that("test Vcmax25 calibration routine p-model (GenSA, rmse)", {
  skip_on_cran()
  drivers <- p_model_drivers_vcmax25
  obs <- rsofun::p_model_validation_vcmax25
  params_fix <- list(
    kphio              = 0.04998, # setup ORG in Stocker et al. 2020 GMD
    kphio_par_a        = 0.01,  # set to zero to disable temperature-dependence of kphio, setup ORG in Stocker et al. 2020 GMD
    kphio_par_b        = 1.0,
    soilm_thetastar    = 0.6 * 240,  # to recover old setup with soil moisture stress
    soilm_betao        = 0.01,
    beta_unitcostratio = 146.0,
    rd_to_vcmax        = 0.014, # value from Atkin et al. 2015 for C3 herbaceous
    # tau_acclim         = 30.0,
    kc_jmax            = 0.41
  )
  
  settings <- list(
    method              = "gensa",
    metric              = rsofun::cost_rmse_pmodel,
    control = list(
      maxit = 2
    ),
    par = list(
      tau_acclim = list(lower = 7, upper = 60, init = 30)
    )
  )
  
  pars <- rsofun::calib_sofun(
    drivers = drivers,
    obs = obs,
    settings = settings,
    # arguments for cost function
    targets = 'vcmax25',
    par_fixed = params_fix,
    parallel = TRUE,
    ncores = 2
  )
  
  # test for correctly returned values
  expect_type(pars, "list")
})

test_that("test joint calibration routine p-model (BT, likelihood maximization)", {
  skip_on_cran()
  drivers <- rbind(gpp = rsofun::p_model_drivers, 
                  vcmax25 = rsofun::p_model_drivers_vcmax25)
  obs <- rbind(gpp = rsofun::p_model_validation,
              vcmax25 = rsofun::p_model_validation_vcmax25)
  params_fix <- list(
    # kphio              = 0.04998, # setup ORG in Stocker et al. 2020 GMD
    kphio_par_a        = 0.01,  # set to zero to disable temperature-dependence of kphio, setup ORG in Stocker et al. 2020 GMD
    kphio_par_b        = 1.0,
    soilm_thetastar    = 0.6 * 240,  # to recover old setup with soil moisture stress
    soilm_betao        = 0.01,
    beta_unitcostratio = 146.0,
    rd_to_vcmax        = 0.014, # value from Atkin et al. 2015 for C3 herbaceous
    tau_acclim         = 30.0,
    kc_jmax            = 0.41
  )
  
  settings <- list(
    method              = "bayesiantools",
    metric              = rsofun::cost_likelihood_pmodel,
    control = list(
      sampler = "DEzs",
      settings = list(
        nrChains = 1,
        burnin = 1,
        iterations = 4
      )
    ),
    par = list(
      kphio = list(lower=0.04, upper=0.09, init=0.05),
      err_gpp = list(lower = 0.01, upper = 4, init = 2),
      err_vcmax25 = list(lower = 0.0001, upper = 0.1, init = 0.005)
    )
  )
  
  pars <- rsofun::calib_sofun(
    drivers = drivers,
    obs = obs,
    settings = settings,
    targets = c('gpp', 'vcmax25'),
    par_fixed = params_fix
  )
  
  # test for correctly returned values
  expect_type(pars, "list")
})