R/BDProb.R
In langfob/bdpg: Package For Building and Testing Biodiversity Problem Generators
#===============================================================================
#
# BDProb.R
#
# Class definitions for biodiversity problems.
#
#===============================================================================
#
# History:
#
# 2016 06 09 - BTL
# Starting to try to redefine many of the data structures as classes
# to get a little standardization and some validity checks.
#
#===============================================================================
# Fairly brief, helpful site about OOP in S4:
# http://thomas-cokelaer.info/blog/2013/01/r-language-object-oriented-programming/
#===============================================================================
# Xu_base_params class.
setClass ("Xu_base_params",
representation (
alpha__ = "numeric",
n__num_groups = "numeric",
p__prop_of_links_between_groups = "numeric",
r__density = "numeric"
),
#-----------------------------------------------------------------
prototype (
alpha__ = as.numeric (NA),
n__num_groups = as.numeric (NA),
p__prop_of_links_between_groups = as.numeric (NA),
r__density = as.numeric (NA)
)
)
#===============================================================================
# Xu_bdpg_extended_params class.
setClass ("Xu_bdpg_extended_params",
representation (
alpha___lower_bound = "numeric",
alpha___upper_bound = "numeric",
derive_alpha_from_n__num_groups_and_opt_frac_0.5 = "logical",
use_unif_rand_alpha__ = "logical",
#2018 03 10 - SHOULD NOT BE IN EXTENDED PARAMS SINCE IT'S A BASE PARAM# n__num_groups = "numeric",
n__num_groups_lower_bound = "numeric",
n__num_groups_upper_bound = "numeric",
use_unif_rand_n__num_groups = "logical",
num_independent_nodes_per_group = "numeric",
use_unif_rand_p__prop_of_links_between_groups = "logical",
p__prop_of_links_between_groups_lower_bound = "numeric",
p__prop_of_links_between_groups_upper_bound = "numeric",
# Used to be either "num_nodes_per_group" or "num_dependent_nodes_per_group".
# Not currently used but not ready to get rid of it yet.
# See function compute_target_num_links_between_2_groups_per_round().
base_for_target_num_links_between_2_groups_per_round = "character",
at_least_1_for_target_num_links_between_2_groups_per_round = "logical",
use_unif_rand_r__density = "logical",
r__density_lower_bound = "numeric",
r__density_upper_bound = "numeric",
integerize = "function"
),
prototype (
alpha___lower_bound = as.numeric (NA),
alpha___upper_bound = as.numeric (NA),
derive_alpha_from_n__num_groups_and_opt_frac_0.5 = NA,
use_unif_rand_alpha__ = NA,
#2018 03 10 - SHOULD NOT BE IN EXTENDED PARAMS SINCE IT'S A BASE PARAM# n__num_groups = as.numeric (NA),
n__num_groups_lower_bound = as.numeric (NA),
n__num_groups_upper_bound = as.numeric (NA),
use_unif_rand_n__num_groups = NA,
num_independent_nodes_per_group = 1,
use_unif_rand_p__prop_of_links_between_groups = NA,
p__prop_of_links_between_groups_lower_bound = as.numeric (NA),
p__prop_of_links_between_groups_upper_bound = as.numeric (NA),
base_for_target_num_links_between_2_groups_per_round = as.character (NA),
at_least_1_for_target_num_links_between_2_groups_per_round = NA,
use_unif_rand_r__density = NA,
r__density_lower_bound = as.numeric (NA),
r__density_upper_bound = as.numeric (NA),
integerize = round
)
)
#===============================================================================
# Xu_derived_params class.
setClass ("Xu_derived_params",
representation (
num_nodes_per_group = "numeric",
num_rounds_of_linking_between_groups = "numeric",
target_num_links_between_2_groups_per_round = "numeric",
num_links_within_one_group = "numeric",
tot_num_links_inside_groups = "numeric",
max_possible_num_links_between_groups = "numeric",
max_possible_tot_num_links = "numeric",
max_possible_tot_num_node_link_pairs = "numeric",
num_independent_nodes_per_group = "numeric",
num_independent_set_nodes = "numeric",
tot_num_nodes = "numeric",
num_dependent_set_nodes = "numeric",
opt_solution_as_frac_of_tot_num_nodes = "numeric" #,
# base_for_target_num_links_between_2_groups_per_round = "numeric",
# at_least_1_for_target_num_links_between_2_groups_per_round = "numeric"
),
#-----------------------------------------------------------------
prototype (
num_nodes_per_group = as.numeric (NA),
num_rounds_of_linking_between_groups = as.numeric (NA),
target_num_links_between_2_groups_per_round = as.numeric (NA),
num_links_within_one_group = as.numeric (NA),
tot_num_links_inside_groups = as.numeric (NA),
max_possible_num_links_between_groups = as.numeric (NA),
max_possible_tot_num_links = as.numeric (NA),
max_possible_tot_num_node_link_pairs = as.numeric (NA),
num_independent_nodes_per_group = as.numeric (NA),
num_independent_set_nodes = as.numeric (NA),
tot_num_nodes = as.numeric (NA),
num_dependent_set_nodes = as.numeric (NA),
opt_solution_as_frac_of_tot_num_nodes = as.numeric (NA) #,
# base_for_target_num_links_between_2_groups_per_round = as.character (NA),
# at_least_1_for_target_num_links_between_2_groups_per_round = as.character (NA)
)
)
#===============================================================================
# Xu_params class.
setClass ("Xu_params",
representation (
base_params = "Xu_base_params",
bdpg_extended_params = "Xu_bdpg_extended_params",
derived_params = "Xu_derived_params"
)
)
#===============================================================================
# PU_spp_pair_info class.
# SHOULD JUST BE A LIST, NOT BE A CLASS, SINCE IT'S ONLY USED TO BUNDLE A
# BUNCH OF RETURN VALUES? (is that true?)
setClass ("PU_spp_pair_info_class",
representation (
PU_spp_pair_indices = "data.frame",
# Need to make these constant names like Xu_bd_problem does
# AND make sure they have the same values in both classes.
PU_col_name = "character",
spp_col_name = "character",
num_PUs = "numeric",
num_spp = "numeric",
correct_solution_cost = "numeric",
Xu_parameters = "Xu_params", # Xu only
correct_solution_vector_is_known = "logical",
dependent_node_IDs = "vector", # Xu only? or at least rename?
nodes = "data.frame", # Xu only? or at least rename?
PU_costs = "vector",
prob_generator_params_known = "logical" # Doesn't belong in this structure?
),
#-----------------------------------------------------------------
prototype (
# PU_spp_pair_indices = "data.frame",
# Need to make these constant names like Xu_bd_problem does
# AND make sure they have the same values in both classes.
PU_col_name = "PU_ID",
spp_col_name = "spp_ID",
num_PUs = as.numeric (NA),
num_spp = as.numeric (NA),
correct_solution_cost = as.numeric (NA),
# Xu_parameters = "Xu_params", # Xu only
correct_solution_vector_is_known = NA,
# dependent_node_IDs = "vector", # Xu only? or at least rename?
# nodes = "data.frame", # Xu only? or at least rename?
# PU_costs = "vector",
prob_generator_params_known = NA
)
)
#===============================================================================
# Xu_prob_gen_info_class class definition.
setClass ("Xu_prob_gen_info_class",
representation (
read_Xu_problem_from_Xu_file = "logical",
infile_name = "character",
Xu_parameters = "Xu_params"
),
#-----------------------------------------------------------------
prototype (
read_Xu_problem_from_Xu_file = NA,
infile_name = as.character (NA)
# ,
# Xu_parameters = "Xu_params"
)
)
#===============================================================================
# APP_prob_info_class class definition.
setClass ("APP_prob_info_class",
representation (
UUID_of_base_problem_that_has_err_added = "character", # UUID string
cost_error_bound = "numeric", # ret_vals_from_apply_cost_errors$cost_error_bound
realized_median_abs_cost_err_frac = "numeric", # ret_vals_from_apply_cost_errors$realized_median_abs_cost_err_frac
realized_mean_abs_cost_err_frac = "numeric", # ret_vals_from_apply_cost_errors$realized_mean_abs_cost_err_frac
realized_sd_abs_cost_err_frac = "numeric", # ret_vals_from_apply_cost_errors$realized_sd_abs_cost_err_frac
original_FP_const_rate = "numeric", # ret_vals_from_apply_errors$original_FP_const_rate
original_FN_const_rate = "numeric", # ret_vals_from_apply_errors$original_FN_const_rate
match_error_counts = "logical", # ret_vals_from_apply_errors$match_error_counts
FP_const_rate = "numeric", # ret_vals_from_apply_errors$FP_const_rate
FN_const_rate = "numeric", # ret_vals_from_apply_errors$FN_const_rate
realized_FP_rate = "numeric", # ret_vals_from_apply_errors$realized_FP_rate
realized_FN_rate = "numeric", # ret_vals_from_apply_errors$realized_FN_rate
realized_Ftot_rate = "numeric", # ret_vals_from_apply_errors$realized_Ftot_rate
euc_realized_FP_and_cost_in_err_frac = "numeric", # realized_euc_in_errors$euc_realized_FP_and_cost_in_err_frac
euc_realized_FN_and_cost_in_err_frac = "numeric", # realized_euc_in_errors$euc_realized_FN_and_cost_in_err_frac
euc_realized_Ftot_and_cost_in_err_frac = "numeric", # realized_euc_in_errors$euc_realized_Ftot_and_cost_in_err_frac
app_num_spp = "numeric", # ret_vals_from_apply_errors$app_num_spp
app_num_PUs = "numeric" #, # ret_vals_from_apply_errors$app_num_PUs
##FixPUsppPairIndices-2018-02-17## app_PU_spp_pair_indices = "data.frame" # ret_vals_from_apply_errors$app_PU_spp_pair_indices
),
#-----------------------------------------------------------------
prototype (
UUID_of_base_problem_that_has_err_added = as.character (NA), # UUID string
cost_error_bound = 0, #as.numeric (NA), # ret_vals_from_apply_cost_errors$cost_error_bound
realized_median_abs_cost_err_frac = 0, #as.numeric (NA), # ret_vals_from_apply_cost_errors$realized_median_abs_cost_err_frac
realized_mean_abs_cost_err_frac = 0, #as.numeric (NA), # ret_vals_from_apply_cost_errors$realized_mean_abs_cost_err_frac
realized_sd_abs_cost_err_frac = 0, #as.numeric (NA), # ret_vals_from_apply_cost_errors$realized_sd_abs_cost_err_frac
original_FP_const_rate = 0, #as.numeric (NA), # ret_vals_from_apply_errors$original_FP_const_rate
original_FN_const_rate = 0, #as.numeric (NA), # ret_vals_from_apply_errors$original_FN_const_rate
match_error_counts = as.logical (NA), # ret_vals_from_apply_errors$match_error_counts
FP_const_rate = 0, #as.numeric (NA), # ret_vals_from_apply_errors$FP_const_rate
FN_const_rate = 0, #as.numeric (NA), # ret_vals_from_apply_errors$FN_const_rate
realized_FP_rate = 0, #as.numeric (NA), # ret_vals_from_apply_errors$realized_FP_rate
realized_FN_rate = 0, #as.numeric (NA), # ret_vals_from_apply_errors$realized_FN_rate
realized_Ftot_rate = 0, #as.numeric (NA), # ret_vals_from_apply_errors$realized_Ftot_rate
euc_realized_FP_and_cost_in_err_frac = 0, #as.numeric (NA), # realized_euc_in_errors$euc_realized_FP_and_cost_in_err_frac
euc_realized_FN_and_cost_in_err_frac = 0, #as.numeric (NA), # realized_euc_in_errors$euc_realized_FN_and_cost_in_err_frac
euc_realized_Ftot_and_cost_in_err_frac = 0, #as.numeric (NA), # realized_euc_in_errors$euc_realized_Ftot_and_cost_in_err_frac
app_num_spp = as.numeric (NA), # ret_vals_from_apply_errors$app_num_spp
app_num_PUs = as.numeric (NA) #, # ret_vals_from_apply_errors$app_num_PUs
# app_PU_spp_pair_indices = "data.frame" # ret_vals_from_apply_errors$app_PU_spp_pair_indices
)
)
#===============================================================================
# Xu_bd_problem class.
# WHAT ELSE NEEDS TO BE INCLUDED TO BE MORE GENERAL,
# E.G., TO DESCRIBE ILP PROBLEMS AND PROBLEMS WITH
# BOUNDARY LENGTHS, ETC.?
# NEED TO WALK THROUGH THE WHOLE PROCESS AS IF
# USING ILP AND MORE COMPLEX MARXAN (E.G., WITH
# BOUNDARY LENGTHS, SPP WEIGHTS, ETC.) AND
# SIMPLE RICHNESS AND ZONATION TO SEE IF THEY
# WILL WORK OR FORCE CHANGES.
# SHOULD THERE JUST BE A PLANNING UNIT DATA CLASS
# THAT CAN BE SUBCLASSED TO HOLD ALL THE VARIANTS
# LIKE BOUNDARY LENGTHS, ETC WHEN NECESSARY?
# SAME IDEA FOR RESERVE SELECTORS?
# 2017 08 21 - BTL
# Since I don't know what form all of these will
# take and there are issues with backward
# compatibility if the class definintion changes,
# I think that I'll have to handle these kinds of
# changes in the future by deriving a new class,
# e.g., by inheriting from this one - though
# I'm not sure what that will do to wrapped
# problems, which already inherit from this one.
# Might need to use multiple inheritance?
setClass ("Xu_bd_problem",
representation (
#-----------------------
# General information
#-----------------------
UUID = "character", # bd_prob
checksum = "character",
rand_seed = "numeric", # Random seed set at the start of building this object.
R_internal_seed_array = "numeric", # R's internal seed state array resulting from setting the seed at start of building this objec.
obj_type_str = "character", # e.g., "RSprob" - used in building name for file and dir names
cor_or_app_str = "character", # "COR" or "APP" - used in building file and dir names
basic_or_wrapped_or_comb_str = "character", # "Base" or "Wrap" or "Comb" - used in building file and dir names
file_name_prefix = "character", # string to combine with uuid to build file and dir names
prob_is_ok = "logical", # bd_prob
#---------------------------------
# Problem generator information
#---------------------------------
prob_generator_params_known = "logical", # bd_prob [COR and APP values? i.e., ERROR MODEL params too? Or, just change this cor_prob_gen_params_known?]
prob_type = "character", # Add this to allow having a pointer to problem-specific information in prob_gen_info?
prob_gen_info = "ANY", # Add this to point to Xu-specific info or something else if different problem source or generator?
# Can't find exactly what to put as the value for a slot whose type you don't want to specify, but on page:
# https://stat.ethz.ch/R-manual/R-patched/library/methods/html/setClass.html
# in the slots section, it mentions "ANY" in a backhanded sort of way, so I'm going to try it here...
#--------------------------------
# Apparent problem information
#--------------------------------
APP_prob_info = "ANY",
combined_err_label = "character",
#-------------------------------------------------------
# Constant naming information for df cols and dirs
# These are the same in all of these problem objects,
# but since this is research code that changes a lot,
# it seems good to have these names saved somewhere
# as self-documentation for the object in case the
# names change in some different version of the code.
# This way, we will always know for a given object,
# what names it assumed when it was built (even
# though this wastes some space by duplicating
# identical values in many objects).
#-------------------------------------------------------
PU_col_name = "character", # bd_prob
spp_col_name = "character", # bd_prob
presences_col_name = "character", # bd_prob
plot_output_dir = "character",
network_output_dir = "character",
bipartite_metrics_file_name_stem = "character",
igraph_metrics_file_name_stem = "character",
#------------------------
# Solution information
#------------------------
# possible problem: "cost" may need to be "EF_value", since cost is going to be something other than just sum of pu costs when using boundary lengths in EF, etc. Maybe not though, since there's always the financial cost, which may or may not be the same as the EF. In that case, you might want to have separate entries for EF_cost and Financial_cost.
correct_solution_cost = "numeric", # bd_prob (only known sometimes) - should this not be COR?, i.e., apparents may have what they perceive to be the opt cost
#correct_solution_cost_is_known = "logical", # bd_prob # Add this? Would it ever be used or is it enough to just have NA for dependent_node_IDs/correct_solution_PU_IDs?
correct_solution_vector_is_known = "logical", # bd_prob
dependent_node_IDs = "numeric", # XU only? OR, IS THIS JUST NAMED POORLY AND SHOULD BE "correct_solution_PU_IDs"
# integer vector (sorted?) of IDs of nodes in the dependent set
# CAN BE DERIVED, SO MAKE IT A FUNCTION?
#-----------------------
# Species information
#-----------------------
num_spp = "numeric", # bd_prob # CAN BE DERIVED, SO MAKE IT A FUNCTION?
all_spp_IDs = "numeric", # bd_prob # CAN BE DERIVED, SO MAKE IT A FUNCTION? # Shouldn't this be vector?
# SHOULD EITHER BE A TABLE WITH MORE IN IT OR THAT TABLE SHOULD ALSO EXIST SEPARATELY.
# TABLE SHOULD CONTAIN ONE ROW PER SPECIES AND THE FOLLOWING COLUMNS:
# SPP_ID, SPP_NAME, TOT_SPP_OCCUPANCY, SPP_TARGET, SPP_WEIGHT, SPP_BASE_PROB_UUID, SPP_ID_IN_BASE_PROB, (spp_name_in_base_prob-could be looked up)
# TOT_SPP_OCCUPANCY means SUM OF PUs or abundances, etc. - to use in rank abundance plots.
# The TOT_SPP_OCCUPANCY values would be different for COR and APP.
#-----------------------------
# Planning unit information
#-----------------------------
num_PUs = "numeric", # bd_prob # CAN BE DERIVED, SO MAKE IT A FUNCTION?
all_PU_IDs = "numeric", # bd_prob # CAN BE DERIVED, SO MAKE IT A FUNCTION? # Shouldn't this be vector?
PU_costs = "vector", # of numeric # bd_prob [COR and APP values]
# ??? Xu only, or just in need of renaming ???
nodes = "data.frame", # not sure - XU only or is this just the list of planning units?
# Should be renamed and rebuilt as a general table of PU info, with one row for each PU:
# PU_ID, PU_NAME, NUM_SPP_ON_PU, IN_SOL_SET, PU_AREA, PU_BOUNDARY_LENGTH, PU_WEIGHT, PU_BASE_PROB_UUID, PU_ID_IN_BASE_PROB, (pu_name_in_base_prob-could be looked up), PU_IN_EXTRA_SET, PU_IN_IND_SET, PU_GROUP_ID
# CURRENTLY, COLS ARE: node_ID, group_ID, dependent_set_member
# MIGHT ALSO WANT TO INCLUDE A COLUMN INDICATING WHETHER NODE IS A)SOLUTION B)INDSET C)EXTRA
# HOWEVER, THAT'S NOT A GENERALIZED PROBLEM CHARACTERISTIC; IT'S PART OF BOTH XU AND PROBLEM COMBINING/WRAPPING.
# STILL, ARE WRAPPING AND COMBINING only ALLOWABLE WHEN YOU KNOW THE XU SETUP?
# OR, CAN YOU DO COMBINING AND WRAPPING WHENEVER YOU KNOW A BASE PROBLEM AND ITS SOLUTION SET
# BY JUST NOT STEPPING ON ANY BASE PROBLEM PU OTHER THAN THE SOLUTION SET.
final_link_counts_for_each_node = "data.frame", # not XU only? CREATED LONG AFTER XU GENERATION?
# SHOULD BE RENAMED TO final_spp_cts_for_each_PU ?
# SHOULD BE A COLUMN IN THE PU INFO TABLE INSTEAD OF A SEPARATE DATA FRAME?
# cols are: PU_ID, freq
# data frame with one column of unique PU IDs (both empty and non-empty) and a second column containing the number of species occurring on the corresponding PU
#-------------------------
# Occupancy information
#-------------------------
PU_spp_pair_indices = "data.frame", # bd_prob - but may be different forms [COR and APP values] {DOES THIS ALSO (NEED TO) REFLECT ABUNDANCE/DENSITY/PROBABILITY/LIKELIHOOD(A LA MAXENT)?}
##FixPUsppPairIndices-2018-02-17## cor_PU_spp_pair_indices = "data.frame", # bd_prob - but may be different forms [COR and APP values] {DOES THIS ALSO (NEED TO) REFLECT ABUNDANCE/DENSITY/PROBABILITY/LIKELIHOOD(A LA MAXENT)?}
# THIS SHOULD INCLUDE "ABUNDANCE/DENSITY/PROBABILITY/LIKELIHOOD(A LA MAXENT)" AS ANOTHER COLUMN
# Table has one row for each PU/SPP pair where the spp occurs on the PU (where "occurs" means non-zero value for ABUND... above)
bpm = "matrix", # bd_prob [COR and APP values]
#-------------------------
# Directory information
#-------------------------
RSprob_dir_names = "list",
#---------------------------------------
# Problem attributes/measures/metrics
#---------------------------------------
# Post-generation measures
compute_network_metrics = "logical", # Toggles whether to ignore all other network metric flags or not.
compute_network_metrics_COR_APP_WRAP = "logical",
use_igraph_metrics = "logical",
use_bipartite_metrics = "logical",
bipartite_metrics_to_use = "character",
bipartite_metrics_from_bipartite_package = "data.frame", # bd_prob [COR and APP values]
bipartite_metrics_from_igraph_package_df = "data.frame" # bd_prob [COR and APP values]
),
#-----------------------------------------------------------------
prototype (
UUID = as.character (NA),
checksum = as.character (NA),
obj_type_str = as.character (NA),
cor_or_app_str = as.character (NA),
basic_or_wrapped_or_comb_str = as.character (NA),
file_name_prefix = as.character (NA),
prob_is_ok = NA,
prob_generator_params_known = NA,
prob_type = as.character (NA),
combined_err_label = "Undefined_combined_err",
PU_col_name = "PU_ID",
spp_col_name = "spp_ID",
presences_col_name = "freq",
plot_output_dir = "plots",
network_output_dir = "networks",
bipartite_metrics_file_name_stem = "bipartite_metrics_from_bipartite_pkg_df",
igraph_metrics_file_name_stem = "bipartite_metrics_from_igraph_pkg_df",
correct_solution_cost = as.numeric (NA),
correct_solution_vector_is_known = NA,
dependent_node_IDs = as.numeric (NA),
num_spp = as.numeric (NA),
all_spp_IDs = as.numeric (NA),
num_PUs = as.numeric (NA),
all_PU_IDs = as.numeric (NA),
compute_network_metrics = NA,
compute_network_metrics_COR_APP_WRAP = NA,
use_igraph_metrics = NA,
use_bipartite_metrics = NA,
bipartite_metrics_to_use = as.character (NA)
)
)
# setMethod ("show", "Xu_bd_problem",
# function (object)
# {
# cat ("Xu_bd_problem = ",
# " \n ")
# }
# )
#===============================================================================
# Xu_wrapped_bd_problem class.
setClass ("Xu_wrapped_bd_problem",
representation (
UUID_of_base_problem_that_is_wrapped = "character", # UUID string
allow_imperfect_wrap = "logical",
wrap_is_imperfect = "logical"
),
#-----------------------------------------------------------------
prototype (
UUID = as.character (NA),
allow_imperfect_wrap = as.logical (NA),
wrap_is_imperfect = as.logical (NA)
),
#-----------------------------------------------------------------
contains = "Xu_bd_problem"
)
#===============================================================================
# EXPdirs class.
setClass ("EXPdirs",
representation (
UUID = "character",
checksum = "character",
# rand_seed = "numeric", # Random seed set at the start of building this object.
# R_internal_seed_array = "numeric", # R's internal seed state array resulting from setting the seed at start of building this objec.
# UUID of the problem the reserve selector
# is running on.
run_on_prob_UUID = "character",
obj_type_str = "character", # e.g., "RSrun_" - used in building name for file and dir names
cor_or_app_str = "character", # "COR" or "APP" - used in building file and dir names
basic_or_wrapped_or_comb_str = "character", # "Base" or "Wrap" or "Comb" - used in building file and dir names
# rs_method_name = "character", # e.g., "Marxan_sa" - used in building file and dir names
file_name_prefix = "character", # string to combine with uuid to build file and dir names
input_dir_name = "character",
output_dir_name = "character",
plot_dir_name = "character" #,
# targets = "vector"
),
#-----------------------------------------------------------------
prototype (
UUID = as.character (NA),
checksum = as.character (NA),
run_on_prob_UUID = as.character (NA),
obj_type_str = as.character (NA),
cor_or_app_str = as.character (NA),
basic_or_wrapped_or_comb_str = as.character (NA),
file_name_prefix = as.character (NA),
input_dir_name = "input",
output_dir_name = "output",
plot_dir_name = "plots"
)
)
#===============================================================================
# RSrun class.
setClass ("RSrun",
representation (
rand_seed = "numeric", # Random seed set at the start of building this object.
R_internal_seed_array = "numeric", # R's internal seed state array resulting from setting the seed at start of building this objec.
rs_method_name = "character", # e.g., "Marxan_sa" - used in building file and dir names
targets = "vector"
),
#-----------------------------------------------------------------
contains = "EXPdirs",
#-----------------------------------------------------------------
prototype (
rs_method_name = as.character (NA)
)
)
#===============================================================================
# RSnet class.
setClass ("RSnet",
representation (
net_method_name = "character" # e.g., "igraph or bipart" - used in building file and dir names
),
#-----------------------------------------------------------------
contains = "EXPdirs",
#-----------------------------------------------------------------
prototype (
net_method_name = as.character (NA)
)
)
#===============================================================================
# # RSrun class.
#
# setClass ("RSrun",
# representation (
# UUID = "character",
# checksum = "character",
#
# rand_seed = "numeric", # Random seed set at the start of building this object.
# R_internal_seed_array = "numeric", # R's internal seed state array resulting from setting the seed at start of building this objec.
#
# # UUID of the problem the reserve selector
# # is running on.
#
# run_on_prob_UUID = "character",
#
# obj_type_str = "character", # e.g., "RSrun_" - used in building name for file and dir names
# cor_or_app_str = "character", # "COR" or "APP" - used in building file and dir names
# basic_or_wrapped_or_comb_str = "character", # "Base" or "Wrap" or "Comb" - used in building file and dir names
# rs_method_name = "character", # e.g., "Marxan_sa" - used in building file and dir names
# file_name_prefix = "character", # string to combine with uuid to build file and dir names
#
# input_dir_name = "character",
# output_dir_name = "character",
# plot_dir_name = "character",
#
# targets = "vector"
# ),
#
# #-----------------------------------------------------------------
#
# prototype (
# UUID = as.character (NA),
# checksum = as.character (NA),
#
# run_on_prob_UUID = as.character (NA),
#
# obj_type_str = as.character (NA),
# cor_or_app_str = as.character (NA),
# basic_or_wrapped_or_comb_str = as.character (NA),
# rs_method_name = as.character (NA),
# file_name_prefix = as.character (NA),
#
# input_dir_name = "input",
# output_dir_name = "output",
# plot_dir_name = "plots"
#
# )
#
# )
#===============================================================================
langfob/bdpg documentation built on Dec. 8, 2022, 5:33 a.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
#===============================================================================
#
# BDProb.R
#
# Class definitions for biodiversity problems.
#
#===============================================================================
#
# History:
#
# 2016 06 09 - BTL
# Starting to try to redefine many of the data structures as classes
# to get a little standardization and some validity checks.
#
#===============================================================================
# Fairly brief, helpful site about OOP in S4:
# http://thomas-cokelaer.info/blog/2013/01/r-language-object-oriented-programming/
#===============================================================================
# Xu_base_params class.
setClass ("Xu_base_params",
representation (
alpha__ = "numeric",
n__num_groups = "numeric",
p__prop_of_links_between_groups = "numeric",
r__density = "numeric"
),
#-----------------------------------------------------------------
prototype (
alpha__ = as.numeric (NA),
n__num_groups = as.numeric (NA),
p__prop_of_links_between_groups = as.numeric (NA),
r__density = as.numeric (NA)
)
)
#===============================================================================
# Xu_bdpg_extended_params class.
setClass ("Xu_bdpg_extended_params",
representation (
alpha___lower_bound = "numeric",
alpha___upper_bound = "numeric",
derive_alpha_from_n__num_groups_and_opt_frac_0.5 = "logical",
use_unif_rand_alpha__ = "logical",
#2018 03 10 - SHOULD NOT BE IN EXTENDED PARAMS SINCE IT'S A BASE PARAM# n__num_groups = "numeric",
n__num_groups_lower_bound = "numeric",
n__num_groups_upper_bound = "numeric",
use_unif_rand_n__num_groups = "logical",
num_independent_nodes_per_group = "numeric",
use_unif_rand_p__prop_of_links_between_groups = "logical",
p__prop_of_links_between_groups_lower_bound = "numeric",
p__prop_of_links_between_groups_upper_bound = "numeric",
# Used to be either "num_nodes_per_group" or "num_dependent_nodes_per_group".
# Not currently used but not ready to get rid of it yet.
# See function compute_target_num_links_between_2_groups_per_round().
base_for_target_num_links_between_2_groups_per_round = "character",
at_least_1_for_target_num_links_between_2_groups_per_round = "logical",
use_unif_rand_r__density = "logical",
r__density_lower_bound = "numeric",
r__density_upper_bound = "numeric",
integerize = "function"
),
prototype (
alpha___lower_bound = as.numeric (NA),
alpha___upper_bound = as.numeric (NA),
derive_alpha_from_n__num_groups_and_opt_frac_0.5 = NA,
use_unif_rand_alpha__ = NA,
#2018 03 10 - SHOULD NOT BE IN EXTENDED PARAMS SINCE IT'S A BASE PARAM# n__num_groups = as.numeric (NA),
n__num_groups_lower_bound = as.numeric (NA),
n__num_groups_upper_bound = as.numeric (NA),
use_unif_rand_n__num_groups = NA,
num_independent_nodes_per_group = 1,
use_unif_rand_p__prop_of_links_between_groups = NA,
p__prop_of_links_between_groups_lower_bound = as.numeric (NA),
p__prop_of_links_between_groups_upper_bound = as.numeric (NA),
base_for_target_num_links_between_2_groups_per_round = as.character (NA),
at_least_1_for_target_num_links_between_2_groups_per_round = NA,
use_unif_rand_r__density = NA,
r__density_lower_bound = as.numeric (NA),
r__density_upper_bound = as.numeric (NA),
integerize = round
)
)
#===============================================================================
# Xu_derived_params class.
setClass ("Xu_derived_params",
representation (
num_nodes_per_group = "numeric",
num_rounds_of_linking_between_groups = "numeric",
target_num_links_between_2_groups_per_round = "numeric",
num_links_within_one_group = "numeric",
tot_num_links_inside_groups = "numeric",
max_possible_num_links_between_groups = "numeric",
max_possible_tot_num_links = "numeric",
max_possible_tot_num_node_link_pairs = "numeric",
num_independent_nodes_per_group = "numeric",
num_independent_set_nodes = "numeric",
tot_num_nodes = "numeric",
num_dependent_set_nodes = "numeric",
opt_solution_as_frac_of_tot_num_nodes = "numeric" #,
# base_for_target_num_links_between_2_groups_per_round = "numeric",
# at_least_1_for_target_num_links_between_2_groups_per_round = "numeric"
),
#-----------------------------------------------------------------
prototype (
num_nodes_per_group = as.numeric (NA),
num_rounds_of_linking_between_groups = as.numeric (NA),
target_num_links_between_2_groups_per_round = as.numeric (NA),
num_links_within_one_group = as.numeric (NA),
tot_num_links_inside_groups = as.numeric (NA),
max_possible_num_links_between_groups = as.numeric (NA),
max_possible_tot_num_links = as.numeric (NA),
max_possible_tot_num_node_link_pairs = as.numeric (NA),
num_independent_nodes_per_group = as.numeric (NA),
num_independent_set_nodes = as.numeric (NA),
tot_num_nodes = as.numeric (NA),
num_dependent_set_nodes = as.numeric (NA),
opt_solution_as_frac_of_tot_num_nodes = as.numeric (NA) #,
# base_for_target_num_links_between_2_groups_per_round = as.character (NA),
# at_least_1_for_target_num_links_between_2_groups_per_round = as.character (NA)
)
)
#===============================================================================
# Xu_params class.
setClass ("Xu_params",
representation (
base_params = "Xu_base_params",
bdpg_extended_params = "Xu_bdpg_extended_params",
derived_params = "Xu_derived_params"
)
)
#===============================================================================
# PU_spp_pair_info class.
# SHOULD JUST BE A LIST, NOT BE A CLASS, SINCE IT'S ONLY USED TO BUNDLE A
# BUNCH OF RETURN VALUES? (is that true?)
setClass ("PU_spp_pair_info_class",
representation (
PU_spp_pair_indices = "data.frame",
# Need to make these constant names like Xu_bd_problem does
# AND make sure they have the same values in both classes.
PU_col_name = "character",
spp_col_name = "character",
num_PUs = "numeric",
num_spp = "numeric",
correct_solution_cost = "numeric",
Xu_parameters = "Xu_params", # Xu only
correct_solution_vector_is_known = "logical",
dependent_node_IDs = "vector", # Xu only? or at least rename?
nodes = "data.frame", # Xu only? or at least rename?
PU_costs = "vector",
prob_generator_params_known = "logical" # Doesn't belong in this structure?
),
#-----------------------------------------------------------------
prototype (
# PU_spp_pair_indices = "data.frame",
# Need to make these constant names like Xu_bd_problem does
# AND make sure they have the same values in both classes.
PU_col_name = "PU_ID",
spp_col_name = "spp_ID",
num_PUs = as.numeric (NA),
num_spp = as.numeric (NA),
correct_solution_cost = as.numeric (NA),
# Xu_parameters = "Xu_params", # Xu only
correct_solution_vector_is_known = NA,
# dependent_node_IDs = "vector", # Xu only? or at least rename?
# nodes = "data.frame", # Xu only? or at least rename?
# PU_costs = "vector",
prob_generator_params_known = NA
)
)
#===============================================================================
# Xu_prob_gen_info_class class definition.
setClass ("Xu_prob_gen_info_class",
representation (
read_Xu_problem_from_Xu_file = "logical",
infile_name = "character",
Xu_parameters = "Xu_params"
),
#-----------------------------------------------------------------
prototype (
read_Xu_problem_from_Xu_file = NA,
infile_name = as.character (NA)
# ,
# Xu_parameters = "Xu_params"
)
)
#===============================================================================
# APP_prob_info_class class definition.
setClass ("APP_prob_info_class",
representation (
UUID_of_base_problem_that_has_err_added = "character", # UUID string
cost_error_bound = "numeric", # ret_vals_from_apply_cost_errors$cost_error_bound
realized_median_abs_cost_err_frac = "numeric", # ret_vals_from_apply_cost_errors$realized_median_abs_cost_err_frac
realized_mean_abs_cost_err_frac = "numeric", # ret_vals_from_apply_cost_errors$realized_mean_abs_cost_err_frac
realized_sd_abs_cost_err_frac = "numeric", # ret_vals_from_apply_cost_errors$realized_sd_abs_cost_err_frac
original_FP_const_rate = "numeric", # ret_vals_from_apply_errors$original_FP_const_rate
original_FN_const_rate = "numeric", # ret_vals_from_apply_errors$original_FN_const_rate
match_error_counts = "logical", # ret_vals_from_apply_errors$match_error_counts
FP_const_rate = "numeric", # ret_vals_from_apply_errors$FP_const_rate
FN_const_rate = "numeric", # ret_vals_from_apply_errors$FN_const_rate
realized_FP_rate = "numeric", # ret_vals_from_apply_errors$realized_FP_rate
realized_FN_rate = "numeric", # ret_vals_from_apply_errors$realized_FN_rate
realized_Ftot_rate = "numeric", # ret_vals_from_apply_errors$realized_Ftot_rate
euc_realized_FP_and_cost_in_err_frac = "numeric", # realized_euc_in_errors$euc_realized_FP_and_cost_in_err_frac
euc_realized_FN_and_cost_in_err_frac = "numeric", # realized_euc_in_errors$euc_realized_FN_and_cost_in_err_frac
euc_realized_Ftot_and_cost_in_err_frac = "numeric", # realized_euc_in_errors$euc_realized_Ftot_and_cost_in_err_frac
app_num_spp = "numeric", # ret_vals_from_apply_errors$app_num_spp
app_num_PUs = "numeric" #, # ret_vals_from_apply_errors$app_num_PUs
##FixPUsppPairIndices-2018-02-17## app_PU_spp_pair_indices = "data.frame" # ret_vals_from_apply_errors$app_PU_spp_pair_indices
),
#-----------------------------------------------------------------
prototype (
UUID_of_base_problem_that_has_err_added = as.character (NA), # UUID string
cost_error_bound = 0, #as.numeric (NA), # ret_vals_from_apply_cost_errors$cost_error_bound
realized_median_abs_cost_err_frac = 0, #as.numeric (NA), # ret_vals_from_apply_cost_errors$realized_median_abs_cost_err_frac
realized_mean_abs_cost_err_frac = 0, #as.numeric (NA), # ret_vals_from_apply_cost_errors$realized_mean_abs_cost_err_frac
realized_sd_abs_cost_err_frac = 0, #as.numeric (NA), # ret_vals_from_apply_cost_errors$realized_sd_abs_cost_err_frac
original_FP_const_rate = 0, #as.numeric (NA), # ret_vals_from_apply_errors$original_FP_const_rate
original_FN_const_rate = 0, #as.numeric (NA), # ret_vals_from_apply_errors$original_FN_const_rate
match_error_counts = as.logical (NA), # ret_vals_from_apply_errors$match_error_counts
FP_const_rate = 0, #as.numeric (NA), # ret_vals_from_apply_errors$FP_const_rate
FN_const_rate = 0, #as.numeric (NA), # ret_vals_from_apply_errors$FN_const_rate
realized_FP_rate = 0, #as.numeric (NA), # ret_vals_from_apply_errors$realized_FP_rate
realized_FN_rate = 0, #as.numeric (NA), # ret_vals_from_apply_errors$realized_FN_rate
realized_Ftot_rate = 0, #as.numeric (NA), # ret_vals_from_apply_errors$realized_Ftot_rate
euc_realized_FP_and_cost_in_err_frac = 0, #as.numeric (NA), # realized_euc_in_errors$euc_realized_FP_and_cost_in_err_frac
euc_realized_FN_and_cost_in_err_frac = 0, #as.numeric (NA), # realized_euc_in_errors$euc_realized_FN_and_cost_in_err_frac
euc_realized_Ftot_and_cost_in_err_frac = 0, #as.numeric (NA), # realized_euc_in_errors$euc_realized_Ftot_and_cost_in_err_frac
app_num_spp = as.numeric (NA), # ret_vals_from_apply_errors$app_num_spp
app_num_PUs = as.numeric (NA) #, # ret_vals_from_apply_errors$app_num_PUs
# app_PU_spp_pair_indices = "data.frame" # ret_vals_from_apply_errors$app_PU_spp_pair_indices
)
)
#===============================================================================
# Xu_bd_problem class.
# WHAT ELSE NEEDS TO BE INCLUDED TO BE MORE GENERAL,
# E.G., TO DESCRIBE ILP PROBLEMS AND PROBLEMS WITH
# BOUNDARY LENGTHS, ETC.?
# NEED TO WALK THROUGH THE WHOLE PROCESS AS IF
# USING ILP AND MORE COMPLEX MARXAN (E.G., WITH
# BOUNDARY LENGTHS, SPP WEIGHTS, ETC.) AND
# SIMPLE RICHNESS AND ZONATION TO SEE IF THEY
# WILL WORK OR FORCE CHANGES.
# SHOULD THERE JUST BE A PLANNING UNIT DATA CLASS
# THAT CAN BE SUBCLASSED TO HOLD ALL THE VARIANTS
# LIKE BOUNDARY LENGTHS, ETC WHEN NECESSARY?
# SAME IDEA FOR RESERVE SELECTORS?
# 2017 08 21 - BTL
# Since I don't know what form all of these will
# take and there are issues with backward
# compatibility if the class definintion changes,
# I think that I'll have to handle these kinds of
# changes in the future by deriving a new class,
# e.g., by inheriting from this one - though
# I'm not sure what that will do to wrapped
# problems, which already inherit from this one.
# Might need to use multiple inheritance?
setClass ("Xu_bd_problem",
representation (
#-----------------------
# General information
#-----------------------
UUID = "character", # bd_prob
checksum = "character",
rand_seed = "numeric", # Random seed set at the start of building this object.
R_internal_seed_array = "numeric", # R's internal seed state array resulting from setting the seed at start of building this objec.
obj_type_str = "character", # e.g., "RSprob" - used in building name for file and dir names
cor_or_app_str = "character", # "COR" or "APP" - used in building file and dir names
basic_or_wrapped_or_comb_str = "character", # "Base" or "Wrap" or "Comb" - used in building file and dir names
file_name_prefix = "character", # string to combine with uuid to build file and dir names
prob_is_ok = "logical", # bd_prob
#---------------------------------
# Problem generator information
#---------------------------------
prob_generator_params_known = "logical", # bd_prob [COR and APP values? i.e., ERROR MODEL params too? Or, just change this cor_prob_gen_params_known?]
prob_type = "character", # Add this to allow having a pointer to problem-specific information in prob_gen_info?
prob_gen_info = "ANY", # Add this to point to Xu-specific info or something else if different problem source or generator?
# Can't find exactly what to put as the value for a slot whose type you don't want to specify, but on page:
# https://stat.ethz.ch/R-manual/R-patched/library/methods/html/setClass.html
# in the slots section, it mentions "ANY" in a backhanded sort of way, so I'm going to try it here...
#--------------------------------
# Apparent problem information
#--------------------------------
APP_prob_info = "ANY",
combined_err_label = "character",
#-------------------------------------------------------
# Constant naming information for df cols and dirs
# These are the same in all of these problem objects,
# but since this is research code that changes a lot,
# it seems good to have these names saved somewhere
# as self-documentation for the object in case the
# names change in some different version of the code.
# This way, we will always know for a given object,
# what names it assumed when it was built (even
# though this wastes some space by duplicating
# identical values in many objects).
#-------------------------------------------------------
PU_col_name = "character", # bd_prob
spp_col_name = "character", # bd_prob
presences_col_name = "character", # bd_prob
plot_output_dir = "character",
network_output_dir = "character",
bipartite_metrics_file_name_stem = "character",
igraph_metrics_file_name_stem = "character",
#------------------------
# Solution information
#------------------------
# possible problem: "cost" may need to be "EF_value", since cost is going to be something other than just sum of pu costs when using boundary lengths in EF, etc. Maybe not though, since there's always the financial cost, which may or may not be the same as the EF. In that case, you might want to have separate entries for EF_cost and Financial_cost.
correct_solution_cost = "numeric", # bd_prob (only known sometimes) - should this not be COR?, i.e., apparents may have what they perceive to be the opt cost
#correct_solution_cost_is_known = "logical", # bd_prob # Add this? Would it ever be used or is it enough to just have NA for dependent_node_IDs/correct_solution_PU_IDs?
correct_solution_vector_is_known = "logical", # bd_prob
dependent_node_IDs = "numeric", # XU only? OR, IS THIS JUST NAMED POORLY AND SHOULD BE "correct_solution_PU_IDs"
# integer vector (sorted?) of IDs of nodes in the dependent set
# CAN BE DERIVED, SO MAKE IT A FUNCTION?
#-----------------------
# Species information
#-----------------------
num_spp = "numeric", # bd_prob # CAN BE DERIVED, SO MAKE IT A FUNCTION?
all_spp_IDs = "numeric", # bd_prob # CAN BE DERIVED, SO MAKE IT A FUNCTION? # Shouldn't this be vector?
# SHOULD EITHER BE A TABLE WITH MORE IN IT OR THAT TABLE SHOULD ALSO EXIST SEPARATELY.
# TABLE SHOULD CONTAIN ONE ROW PER SPECIES AND THE FOLLOWING COLUMNS:
# SPP_ID, SPP_NAME, TOT_SPP_OCCUPANCY, SPP_TARGET, SPP_WEIGHT, SPP_BASE_PROB_UUID, SPP_ID_IN_BASE_PROB, (spp_name_in_base_prob-could be looked up)
# TOT_SPP_OCCUPANCY means SUM OF PUs or abundances, etc. - to use in rank abundance plots.
# The TOT_SPP_OCCUPANCY values would be different for COR and APP.
#-----------------------------
# Planning unit information
#-----------------------------
num_PUs = "numeric", # bd_prob # CAN BE DERIVED, SO MAKE IT A FUNCTION?
all_PU_IDs = "numeric", # bd_prob # CAN BE DERIVED, SO MAKE IT A FUNCTION? # Shouldn't this be vector?
PU_costs = "vector", # of numeric # bd_prob [COR and APP values]
# ??? Xu only, or just in need of renaming ???
nodes = "data.frame", # not sure - XU only or is this just the list of planning units?
# Should be renamed and rebuilt as a general table of PU info, with one row for each PU:
# PU_ID, PU_NAME, NUM_SPP_ON_PU, IN_SOL_SET, PU_AREA, PU_BOUNDARY_LENGTH, PU_WEIGHT, PU_BASE_PROB_UUID, PU_ID_IN_BASE_PROB, (pu_name_in_base_prob-could be looked up), PU_IN_EXTRA_SET, PU_IN_IND_SET, PU_GROUP_ID
# CURRENTLY, COLS ARE: node_ID, group_ID, dependent_set_member
# MIGHT ALSO WANT TO INCLUDE A COLUMN INDICATING WHETHER NODE IS A)SOLUTION B)INDSET C)EXTRA
# HOWEVER, THAT'S NOT A GENERALIZED PROBLEM CHARACTERISTIC; IT'S PART OF BOTH XU AND PROBLEM COMBINING/WRAPPING.
# STILL, ARE WRAPPING AND COMBINING only ALLOWABLE WHEN YOU KNOW THE XU SETUP?
# OR, CAN YOU DO COMBINING AND WRAPPING WHENEVER YOU KNOW A BASE PROBLEM AND ITS SOLUTION SET
# BY JUST NOT STEPPING ON ANY BASE PROBLEM PU OTHER THAN THE SOLUTION SET.
final_link_counts_for_each_node = "data.frame", # not XU only? CREATED LONG AFTER XU GENERATION?
# SHOULD BE RENAMED TO final_spp_cts_for_each_PU ?
# SHOULD BE A COLUMN IN THE PU INFO TABLE INSTEAD OF A SEPARATE DATA FRAME?
# cols are: PU_ID, freq
# data frame with one column of unique PU IDs (both empty and non-empty) and a second column containing the number of species occurring on the corresponding PU
#-------------------------
# Occupancy information
#-------------------------
PU_spp_pair_indices = "data.frame", # bd_prob - but may be different forms [COR and APP values] {DOES THIS ALSO (NEED TO) REFLECT ABUNDANCE/DENSITY/PROBABILITY/LIKELIHOOD(A LA MAXENT)?}
##FixPUsppPairIndices-2018-02-17## cor_PU_spp_pair_indices = "data.frame", # bd_prob - but may be different forms [COR and APP values] {DOES THIS ALSO (NEED TO) REFLECT ABUNDANCE/DENSITY/PROBABILITY/LIKELIHOOD(A LA MAXENT)?}
# THIS SHOULD INCLUDE "ABUNDANCE/DENSITY/PROBABILITY/LIKELIHOOD(A LA MAXENT)" AS ANOTHER COLUMN
# Table has one row for each PU/SPP pair where the spp occurs on the PU (where "occurs" means non-zero value for ABUND... above)
bpm = "matrix", # bd_prob [COR and APP values]
#-------------------------
# Directory information
#-------------------------
RSprob_dir_names = "list",
#---------------------------------------
# Problem attributes/measures/metrics
#---------------------------------------
# Post-generation measures
compute_network_metrics = "logical", # Toggles whether to ignore all other network metric flags or not.
compute_network_metrics_COR_APP_WRAP = "logical",
use_igraph_metrics = "logical",
use_bipartite_metrics = "logical",
bipartite_metrics_to_use = "character",
bipartite_metrics_from_bipartite_package = "data.frame", # bd_prob [COR and APP values]
bipartite_metrics_from_igraph_package_df = "data.frame" # bd_prob [COR and APP values]
),
#-----------------------------------------------------------------
prototype (
UUID = as.character (NA),
checksum = as.character (NA),
obj_type_str = as.character (NA),
cor_or_app_str = as.character (NA),
basic_or_wrapped_or_comb_str = as.character (NA),
file_name_prefix = as.character (NA),
prob_is_ok = NA,
prob_generator_params_known = NA,
prob_type = as.character (NA),
combined_err_label = "Undefined_combined_err",
PU_col_name = "PU_ID",
spp_col_name = "spp_ID",
presences_col_name = "freq",
plot_output_dir = "plots",
network_output_dir = "networks",
bipartite_metrics_file_name_stem = "bipartite_metrics_from_bipartite_pkg_df",
igraph_metrics_file_name_stem = "bipartite_metrics_from_igraph_pkg_df",
correct_solution_cost = as.numeric (NA),
correct_solution_vector_is_known = NA,
dependent_node_IDs = as.numeric (NA),
num_spp = as.numeric (NA),
all_spp_IDs = as.numeric (NA),
num_PUs = as.numeric (NA),
all_PU_IDs = as.numeric (NA),
compute_network_metrics = NA,
compute_network_metrics_COR_APP_WRAP = NA,
use_igraph_metrics = NA,
use_bipartite_metrics = NA,
bipartite_metrics_to_use = as.character (NA)
)
)
# setMethod ("show", "Xu_bd_problem",
# function (object)
# {
# cat ("Xu_bd_problem = ",
# " \n ")
# }
# )
#===============================================================================
# Xu_wrapped_bd_problem class.
setClass ("Xu_wrapped_bd_problem",
representation (
UUID_of_base_problem_that_is_wrapped = "character", # UUID string
allow_imperfect_wrap = "logical",
wrap_is_imperfect = "logical"
),
#-----------------------------------------------------------------
prototype (
UUID = as.character (NA),
allow_imperfect_wrap = as.logical (NA),
wrap_is_imperfect = as.logical (NA)
),
#-----------------------------------------------------------------
contains = "Xu_bd_problem"
)
#===============================================================================
# EXPdirs class.
setClass ("EXPdirs",
representation (
UUID = "character",
checksum = "character",
# rand_seed = "numeric", # Random seed set at the start of building this object.
# R_internal_seed_array = "numeric", # R's internal seed state array resulting from setting the seed at start of building this objec.
# UUID of the problem the reserve selector
# is running on.
run_on_prob_UUID = "character",
obj_type_str = "character", # e.g., "RSrun_" - used in building name for file and dir names
cor_or_app_str = "character", # "COR" or "APP" - used in building file and dir names
basic_or_wrapped_or_comb_str = "character", # "Base" or "Wrap" or "Comb" - used in building file and dir names
# rs_method_name = "character", # e.g., "Marxan_sa" - used in building file and dir names
file_name_prefix = "character", # string to combine with uuid to build file and dir names
input_dir_name = "character",
output_dir_name = "character",
plot_dir_name = "character" #,
# targets = "vector"
),
#-----------------------------------------------------------------
prototype (
UUID = as.character (NA),
checksum = as.character (NA),
run_on_prob_UUID = as.character (NA),
obj_type_str = as.character (NA),
cor_or_app_str = as.character (NA),
basic_or_wrapped_or_comb_str = as.character (NA),
file_name_prefix = as.character (NA),
input_dir_name = "input",
output_dir_name = "output",
plot_dir_name = "plots"
)
)
#===============================================================================
# RSrun class.
setClass ("RSrun",
representation (
rand_seed = "numeric", # Random seed set at the start of building this object.
R_internal_seed_array = "numeric", # R's internal seed state array resulting from setting the seed at start of building this objec.
rs_method_name = "character", # e.g., "Marxan_sa" - used in building file and dir names
targets = "vector"
),
#-----------------------------------------------------------------
contains = "EXPdirs",
#-----------------------------------------------------------------
prototype (
rs_method_name = as.character (NA)
)
)
#===============================================================================
# RSnet class.
setClass ("RSnet",
representation (
net_method_name = "character" # e.g., "igraph or bipart" - used in building file and dir names
),
#-----------------------------------------------------------------
contains = "EXPdirs",
#-----------------------------------------------------------------
prototype (
net_method_name = as.character (NA)
)
)
#===============================================================================
# # RSrun class.
#
# setClass ("RSrun",
# representation (
# UUID = "character",
# checksum = "character",
#
# rand_seed = "numeric", # Random seed set at the start of building this object.
# R_internal_seed_array = "numeric", # R's internal seed state array resulting from setting the seed at start of building this objec.
#
# # UUID of the problem the reserve selector
# # is running on.
#
# run_on_prob_UUID = "character",
#
# obj_type_str = "character", # e.g., "RSrun_" - used in building name for file and dir names
# cor_or_app_str = "character", # "COR" or "APP" - used in building file and dir names
# basic_or_wrapped_or_comb_str = "character", # "Base" or "Wrap" or "Comb" - used in building file and dir names
# rs_method_name = "character", # e.g., "Marxan_sa" - used in building file and dir names
# file_name_prefix = "character", # string to combine with uuid to build file and dir names
#
# input_dir_name = "character",
# output_dir_name = "character",
# plot_dir_name = "character",
#
# targets = "vector"
# ),
#
# #-----------------------------------------------------------------
#
# prototype (
# UUID = as.character (NA),
# checksum = as.character (NA),
#
# run_on_prob_UUID = as.character (NA),
#
# obj_type_str = as.character (NA),
# cor_or_app_str = as.character (NA),
# basic_or_wrapped_or_comb_str = as.character (NA),
# rs_method_name = as.character (NA),
# file_name_prefix = as.character (NA),
#
# input_dir_name = "input",
# output_dir_name = "output",
# plot_dir_name = "plots"
#
# )
#
# )
#===============================================================================
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.