Merge pull request #45 from ncn-foreigners/dev

Dev

DESCRIPTION

@@ -1,6 +1,6 @@
 Package: nonprobsvy
 Type: Package
-Title: Package for Inference Based on Non-Probability Samples
+Title: Inference Based on Non-Probability Samples
 Version: 0.1.0
 Authors@R:
     c(person(given = "Łukasz",
@@ -16,7 +16,7 @@ Authors@R:
              family = "Chlebicki",
              role = "ctb",
              email = "[email protected]"))
-Description: An R package for statistical inference with non-probability samples when auxiliary information from external sources such as probability samples or population totals or means is available. Details can be found in: Wu et al. (2020) <doi:10.1080/01621459.2019.1677241>, Kim et al. (2021) <doi:10.1111/rssa.12696>, Wu et al. (2023) <https://www150.statcan.gc.ca/n1/pub/12-001-x/2022002/article/00002-eng.htm>, Kim et al. (2021) <https://www150.statcan.gc.ca/n1/pub/12-001-x/2021001/article/00004-eng.htm>, Kim et al. (2020) <doi:10.1111/rssb.12354>.
+Description: Statistical inference with non-probability samples when auxiliary information from external sources such as probability samples or population totals or means is available. Details can be found in: Wu et al. (2020) <doi:10.1080/01621459.2019.1677241>, Kim et al. (2021) <doi:10.1111/rssa.12696>, Wu et al. (2023) <https://www150.statcan.gc.ca/n1/pub/12-001-x/2022002/article/00002-eng.htm>, Kim et al. (2021) <https://www150.statcan.gc.ca/n1/pub/12-001-x/2021001/article/00004-eng.htm>, Kim et al. (2020) <doi:10.1111/rssb.12354>.
 License: MIT + file LICENSE
 Encoding: UTF-8
 LazyData: true

R/cloglogModel.R

@@ -4,6 +4,8 @@
 #' @description \code{cloglog_model_nonprobsvy} returns all the methods/objects/functions required to estimate the model, assuming a cloglog link function.
 #' @param ... Additional, optional arguments.
 #'
+#' @return List with selected methods/objects/functions.
+#'
 #' @seealso
 #'
 #' [nonprob()] -- for fitting procedure with non-probability samples.

R/pmm.R

@@ -38,7 +38,7 @@ pmm_nonprobsvy <- function(outcome,
         outcome,
         data,
         span = .2,
-        control = stats::loess.control(surface = "direct")
+        control = stats::loess.control(surface = "interpolate", trace.hat = "approximate")
       )
       mm$data <- data
       mm$formula <- outcome
@@ -216,7 +216,7 @@ pmm_exact <- function(pi_ij,
           formula = model_obj$model$glm_object$formula,
           data = model_obj$model$glm_object$data[boot_samp, , drop = FALSE],
           span = .2,
-          control = stats::loess.control(surface = "direct")
+          control = stats::loess.control(surface = "interpolate", trace.hat = "approximate")
         )
       )
       XX <- predict(

R/simple_methods.R

@@ -17,6 +17,7 @@ pop.size.nonprobsvy <- function(object,
 #' @description Estimate size of population
 #' @param object object returned by `nonprobsvy`.
 #' @param ... additional parameters
+#' @return Vector returning the value of the estimated population size.
 #' @export
 pop.size <- function(object, ...) {
   UseMethod("pop.size")

R/summary.R

@@ -10,6 +10,30 @@
 #' @param cov Covariance matrix corresponding to regression parameters
 #' @param ... Additional optional arguments
 #'
+#' @return An object of \code{summary_nonprobsvy} class containing:
+#' \itemize{
+#' \item \code{call} -- A call which created \code{object}.
+#' \item \code{pop_total} -- A list containing information about the estimated population mean, its standard error and confidence interval.
+#' \item \code{sample_size} -- The size of the samples used in the model.
+#' \item \code{population_size} -- The estimated size of the population from which the nonoprobability sample was drawn.
+#' \item \code{test} -- Type of statistical test performed.
+#' \item \code{control} -- A List of control parameters used in fitting the model.
+#' \item \code{model} -- A descriptive name of the model used, e.g., "Doubly-Robust", "Inverse probability weighted", or "Mass Imputation".
+#' \item \code{aic} -- Akaike's information criterion.
+#' \item \code{bic} -- Bayesian (Schwarz's) information criterion.
+#' \item \code{residuals} -- Residuals from the model, providing information on the difference between observed and predicted values.
+#' \item \code{likelihood} -- Logarithm of likelihood function evaluated at coefficients.
+#' \item \code{df_residual} -- Residual degrees of freedom.
+#' \item \code{weights} -- Distribution of estimated weights obtained from the model.
+#' \item \code{coef} -- Regression coefficients estimated by the model.
+#' \item \code{std_err} -- Standard errors of the regression coefficients.
+#' \item \code{w_val} -- Wald statistic values for the significance testing of coefficients.
+#' \item \code{p_values} -- P-values corresponding to the Wald statistic values, assessing the significance of coefficients.
+#' \item \code{crr} -- The correlation matrix of the model coefficients, if requested.
+#' \item \code{confidence_interval_coef} -- Confidence intervals for the model coefficients.
+#' \item \code{names} -- Names of the fitted models.
+#' }
+#'
 #'
 #' @method summary nonprobsvy
 #' @importFrom stats pt

src/Makevars

@@ -13,7 +13,7 @@
 ## _In general_ we should no longer need to set a standard as any recent R
 ## installation will do the right thing. Should you need it, uncomment it and
 ## set the appropriate value, possibly CXX17.
-#CXX_STD = CXX11
+# CXX_STD = CXX11
 
 PKG_CXXFLAGS = $(SHLIB_OPENMP_CXXFLAGS)
 PKG_LIBS = $(SHLIB_OPENMP_CXXFLAGS) $(LAPACK_LIBS) $(BLAS_LIBS) $(FLIBS)