Calculate the AIC for a column of an mmkin object

Provides a convenient way to compare different kinetic models fitted to the same dataset.

Usage

# S3 method for mmkin
AIC(object, ..., k = 2)

# S3 method for mmkin
BIC(object, ...)

Arguments

object: An object of class mmkin, containing only one column.
...: For compatibility with the generic method
k: As in the generic method

Value

As in the generic method (a numeric value for single fits, or a dataframe if there are several fits in the column).

Author

Johannes Ranke

Examples


   # skip, as it takes > 10 s on winbuilder
  f <- mmkin(c("SFO", "FOMC", "DFOP"),
    list("FOCUS A" = FOCUS_2006_A,
         "FOCUS C" = FOCUS_2006_C), cores = 1, quiet = TRUE)
#> Warning: Optimisation did not converge:
#> false convergence (8)
  # We get a warning because the FOMC model does not converge for the
  # FOCUS A dataset, as it is well described by SFO

  AIC(f["SFO", "FOCUS A"]) # We get a single number for a single fit
#> [1] 55.28197
  AIC(f[["SFO", "FOCUS A"]]) # or when extracting an mkinfit object
#> [1] 55.28197

  # For FOCUS A, the models fit almost equally well, so the higher the number
  # of parameters, the higher (worse) the AIC
  AIC(f[, "FOCUS A"])
#>      df      AIC
#> SFO   3 55.28197
#> FOMC  4 57.28198
#> DFOP  5 59.28197
  AIC(f[, "FOCUS A"], k = 0) # If we do not penalize additional parameters, we get nearly the same
#>      df      AIC
#> SFO   3 49.28197
#> FOMC  4 49.28198
#> DFOP  5 49.28197
  BIC(f[, "FOCUS A"])        # Comparing the BIC gives a very similar picture
#>      df      BIC
#> SFO   3 55.52030
#> FOMC  4 57.59974
#> DFOP  5 59.67918

  # For FOCUS C, the more complex models fit better
  AIC(f[, "FOCUS C"])
#>      df      AIC
#> SFO   3 59.29336
#> FOMC  4 44.68652
#> DFOP  5 29.02372
  BIC(f[, "FOCUS C"])
#>      df      BIC
#> SFO   3 59.88504
#> FOMC  4 45.47542
#> DFOP  5 30.00984