R/approx_evdsi.R
approx_evdsi.RdCalculate the expected value of the management decision given survey information. This metric describes the value of the management decision that is expected when the decision maker conducts a surveys a set of sites to inform the decision. To speed up the calculations, an approximation method is used.
approx_evdsi(
site_data,
feature_data,
site_detection_columns,
site_n_surveys_columns,
site_probability_columns,
site_management_cost_column,
site_survey_scheme_column,
site_survey_cost_column,
feature_survey_column,
feature_survey_sensitivity_column,
feature_survey_specificity_column,
feature_model_sensitivity_column,
feature_model_specificity_column,
feature_target_column,
total_budget,
site_management_locked_in_column = NULL,
site_management_locked_out_column = NULL,
prior_matrix = NULL,
n_approx_replicates = 100,
n_approx_outcomes_per_replicate = 10000,
seed = 500
)sf::sf() object with site data.
base::data.frame() object with feature data.
character names of numeric
columns in the argument to site_data that contain the proportion of
surveys conducted within each site that detected each feature.
Each column should correspond to a different feature, and contain
a proportion value (between zero and one). If a site has
not previously been surveyed, a value of zero should be used.
character names of numeric
columns in the argument to site_data that contain the total
number of surveys conducted for each each feature within each site.
Each column should correspond to a different feature, and contain
a non-negative integer number (e.g. 0, 1, 2, 3). If a site has
not previously been surveyed, a value of zero should be used.
character names of numeric
columns in the argument to site_data that contain modeled
probabilities of occupancy for each feature in each site.
Each column should correspond to a different feature, and contain
probability data (values between zero and one). No missing (NA)
values are permitted in these columns.
character name of column in the
argument to site_data that contains costs for managing each
site for conservation. This column should have numeric values that
are equal to or greater than zero. No missing (NA) values are
permitted in this column.
character name of logical
(TRUE / FALSE) column in the argument to site_data
that indicates which sites are selected in the scheme or not.
No missing NA values are permitted. Additionally, only sites
that are missing data can be selected or surveying (as per the
argument to site_detection_columns).
character name of column in the
argument to site_data that contains costs for surveying each
site. This column should have numeric values that are equal to
or greater than zero. No missing (NA) values are permitted in this
column.
character name of the column in the
argument to feature_data that contains logical (TRUE /
FALSE) values indicating if the feature will be surveyed in
the planned surveys or not. Note that considering additional features will
rapidly increase computational burden, and so it is only recommended to
consider features that are of specific conservation interest.
No missing (NA) values are permitted in this column.
character name of the
column in the argument to feature_data that contains
probability of future surveys correctly detecting a presence of each
feature in a given site (i.e. the sensitivity of the survey methodology).
This column should have numeric values that are between zero and
one. No missing (NA) values are permitted in this column.
character name of the
column in the argument to feature_data that contains
probability of future surveys correctly detecting an absence of each
feature in a given site (i.e. the specificity of the survey methodology).
This column should have numeric values that are between zero and
one. No missing (NA) values are permitted in this column.
character name of the
column in the argument to feature_data that contains
probability of the initial models correctly predicting a presence of each
feature in a given site (i.e. the sensitivity of the models).
This column should have numeric values that are between zero and
one. No missing (NA) values are permitted in this column.
This should ideally be calculated using
fit_xgb_occupancy_models() or
fit_hglm_occupancy_models().
character name of the
column in the argument to feature_data that contains
probability of the initial models correctly predicting an absence of each
feature in a given site (i.e. the specificity of the models).
This column should have numeric values that are between zero and
one. No missing (NA) values are permitted in this column.
This should ideally be calculated using
fit_xgb_occupancy_models() or
fit_hglm_occupancy_models().
character name of the column in the
argument to feature_data that contains the \(target\)
values used to parametrize the conservation benefit of managing of each
feature.
This column should have numeric values that
are equal to or greater than zero. No missing (NA) values are
permitted in this column.
numeric maximum expenditure permitted
for conducting surveys and managing sites for conservation.
character name of the column
in the argument to site_data that contains logical
(TRUE / FALSE) values indicating which sites should
be locked in for (TRUE) being managed for conservation or
(FALSE) not. No missing (NA) values are permitted in this
column. This is useful if some sites have already been earmarked for
conservation, or if some sites are already being managed for conservation.
Defaults to NULL such that no sites are locked in.
character name of the column
in the argument to site_data that contains logical
(TRUE / FALSE) values indicating which sites should
be locked out for (TRUE) being managed for conservation or
(FALSE) not. No missing (NA) values are permitted in this
column. This is useful if some sites could potentially be surveyed
to improve model predictions even if they cannot be managed for
conservation. Defaults to NULL such that no sites are locked out.
numeric matrix containing
the prior probability of each feature occupying each site.
Rows correspond to features, and columns correspond to sites.
Defaults to NULL such that prior data is calculated automatically
using prior_probability_matrix().
integer number of replicates to use for
approximating the expected value calculations. Defaults to 100.
integer number of outcomes to
use per replicate for approximation calculations. Defaults to 10000.
integer state of the random number generator for
simulating outcomes when conducting the value of information analyses.
Defaults to 500.
A numeric vector containing the expected values for each
replicate.
This function uses approximation methods to estimate the
expected value calculations. The accuracy of these
calculations depend on the arguments to
n_approx_replicates and n_approx_outcomes_per_replicate, and
so you may need to increase these parameters for large problems.
# set seeds for reproducibility
set.seed(123)
# load example site data
data(sim_sites)
print(sim_sites)
#> Simple feature collection with 6 features and 13 fields
#> Geometry type: POINT
#> Dimension: XY
#> Bounding box: xmin: 0.10513 ymin: 0.04556193 xmax: 0.9764926 ymax: 0.8637977
#> CRS: NA
#> # A tibble: 6 × 14
#> survey_cost management_cost f1 f2 f3 n1 n2 n3 e1 e2
#> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
#> 1 14 102 1 1 1 3 3 3 1.00 -0.848
#> 2 25 90 0 0 0 0 0 0 -1.44 1.27
#> 3 25 165 1 0.6 0 5 5 5 1.25 0.817
#> 4 17 104 0 0 0 0 0 0 -0.484 -0.292
#> 5 18 100 0 0 0 0 0 0 0.0135 0.380
#> 6 15 94 0 0 0 0 0 0 -0.347 -1.33
#> # ℹ 4 more variables: p1 <dbl>, p2 <dbl>, p3 <dbl>, geometry <POINT>
# load example feature data
data(sim_features)
print(sim_features)
#> # A tibble: 3 × 7
#> name survey survey_sensitivity survey_specificity model_sensitivity
#> <chr> <lgl> <dbl> <dbl> <dbl>
#> 1 f1 TRUE 0.954 0.886 0.718
#> 2 f2 TRUE 0.974 0.875 0.705
#> 3 f3 TRUE 0.956 0.823 0.768
#> # ℹ 2 more variables: model_specificity <dbl>, target <dbl>
# set total budget for managing sites for conservation
# (i.e. 50% of the cost of managing all sites)
total_budget <- sum(sim_sites$management_cost) * 0.5
# create a survey scheme that samples the first two sites that
# are missing data
sim_sites$survey_site <- FALSE
sim_sites$survey_site[which(sim_sites$n1 < 0.5)[1:2]] <- TRUE
# calculate expected value of management decision given the survey
# information using approximation method
approx_ev_survey <- approx_evdsi(
sim_sites, sim_features,
c("f1", "f2", "f3"), c("n1", "n2", "n3"), c("p1", "p2", "p3"),
"management_cost", "survey_site",
"survey_cost", "survey", "survey_sensitivity", "survey_specificity",
"model_sensitivity", "model_specificity",
"target", total_budget)
# print mean value
print(mean(approx_ev_survey))
#> [1] 2.99216