direct_optimise: Maximum likelihood estimate for single frequency count table
In statdivlab/rre: Replicate Richness Estimator

View source: R/penalty_single_sample_util.R

Finds the MLE for a single frequency count table with the option of several search schemes and penalization choices.

direct_optimise(fct, penalty = NULL, lambda = NULL,
  search_scheme = "grid", multiplier = 10, c_seq_len = 100,
  starts = NULL, alpha_min = 1e-10, delta_min = 1e-10,
  alpha_max = 1e+05, delta_max = 1e+05, forced_ccc_lower_bound = NA)

`fct`	a frequency count table, see `make_frequency_count_table` for more information.
`penalty`	method of penalizing/regularizing the maximum likelihood solution.
`lambda`	parameter for the h1 penalization method.
`search_scheme`	method for iterating over candidate C values.
`multiplier`	Richness values from the observed richness to `multiplier`*(observed richness) will be tested (default: 10).
`c_seq_len`	Number of candidate C values we will maximize over.
`starts`	A dataframe or matrix of optimization starts. See details.
`alpha_min, delta_min, alpha_max, delta_max`	box constraints for optimization over `alpha` and `delta`.
`forced_ccc_lower_bound`	allows the user to force a lower bound to the grid search for C.

direct_optimise(nb_fct_simulation(1000, 100, 0.95))

# Suppose we want something quicker:
direct_optimise(nb_fct_simulation(1000, 100, 0.95),
    search_scheme = "bisection",
    c_seq_len = 20)