Skip to content

Homogenize validation data.sets as arguments to cost_likelihood_pmodel() and cost_likelihood_biomee() #236

@fabern

Description

@fabern

After the inputs to the running of the model is homogenized (#228),
also the inputs to the calibration would need to be homogenized.

Some work has already been done to reduce code redundundancy in the likelihood/RMSE functions in PR: #233
Either within that PR or after it, we should also homogenize the validation data sets, i.e. rsofun::biomee_validation, rsofun::p_model_validation_vcmax25, and rsofun::p_model_validation.

# TODO: homogenize format rsofun::biomee_validation with rsofun::p_model_validation_vcmax25
#       for pmodel it is a wide data structure, and for biomee it is a long data structure
# rsofun::p_model_validation |> tidyr::unnest(data)
#  # A tibble: 2,190 × 4
#   sitename date         gpp gpp_unc
#   <chr>    <date>     <dbl>   <dbl>
# 1 FR-Pue   2007-01-01 2.21  0.0108 
# 2 FR-Pue   2007-01-02 2.23  0.00475
# 3 FR-Pue   2007-01-03 2.48  0.00727
# 4 FR-Pue   2007-01-04 1.71  0.00516
# rsofun::p_model_validation_vcmax25 |> tidyr::unnest(data)
# # A tibble: 4 × 3
# # Groups:   sitename [4]
#   sitename               vcmax25 vcmax25_unc
#   <chr>                    <dbl>       <dbl>
# 1 Reichetal_Colorado   0.0000339   0.0000136
# 2 Reichetal_New_Mexico 0.0000757   0.0000163
# 3 Reichetal_Venezuela  0.0000472   0.0000164
# 4 Reichetal_Wisconsin  0.0000502   0.0000147
# rsofun::biomee_validation |> tidyr::unnest(data)
#     sitename variables targets_obs
#   <chr>    <chr>           <dbl>
# 1 CH-Lae   GPP              1.86
# 2 CH-Lae   LAI              6.49
# 3 CH-Lae   Density        296.
# 4 CH-Lae   Biomass         44.5

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type
    No fields configured for issues without a type.

    Projects

    No projects

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions