Updated documentation

R/accuracy.R

@@ -68,18 +68,6 @@ plot.clubproaccuracy <- function(x, ...) {
 }
 
 
-#' Print prediction accuracy.
-#'
-#' @details
-#' Print a table of prediction accuracy by category
-#'
-#' @param x an object of class "clubproaccuracy"
-#' @param ... ignored
-#' @return called for side-effects only
-#' @examples
-#' mod <- club(rate ~ dose, data = caffeine)
-#' z <- accuracy(mod)
-#' print(z)
 #' @export
 print.clubproaccuracy <- function(x, ...) {
     print(unclass(x))

R/classify.R

@@ -160,13 +160,6 @@ club.formula <- function(f, data, imprecision = 0, nreps = 1000L, normalise_cols
 }
 
 
-#' Print the model call.
-#' @param x an object of class "clubprofit"
-#' @param ... ignored
-#' @return No return value, called for side effects.
-#' @examples
-#' mod <- club(rate ~ dose, data = caffeine)
-#' print(mod)
 #' @export
 print.clubprofit <- function(x, ...) {
     print(x$call)

R/compare.R

@@ -46,11 +46,10 @@ compare.clubprofit <- function(m1, m2) {
     pcc_diff <- abs(pcc(m1) - pcc(m2))
     diff_dist <- pcc_replicates(m1) - pcc_replicates(m2)
     cval <- length(abs(diff_dist)[abs(diff_dist) >= pcc_diff]) / m1$nreps
-    structure(
-        list(pcc_diff = pcc_diff,
-             diff_dist = diff_dist,
-             cval = cval),
-        class = "clubprocomparison")
+    structure(list(pcc_diff = pcc_diff,
+                   diff_dist = diff_dist,
+                   cval = cval),
+              class = "clubprocomparison")
 }
 
 
@@ -66,13 +65,11 @@ compare.clubprofit <- function(m1, m2) {
 #' @export
 summary.clubprocomparison <- function(object, ...) {
     structure(
-        list(
-            pcc_diff = object$pcc_diff,
-            min_pcc_difference = min(object$diff_dist),
-            max_pcc_difference = max(object$diff_dist),
-            nreps = length(object$diff_dist),
-            cval = object$cval
-        ),
+        list(pcc_diff = object$pcc_diff,
+             min_pcc_difference = min(object$diff_dist),
+             max_pcc_difference = max(object$diff_dist),
+             nreps = length(object$diff_dist),
+             cval = object$cval),
         class = "summary.clubprocomparison"
     )
 }
@@ -84,7 +81,7 @@ print.summary.clubprocomparison <- function(x, ..., digits = 2L) {
     stopifnot("digits cannot be negative"= digits >= 0)
     stopifnot("digits must be a single number"= length(digits) == 1)
 
-    cat("********** Model Comparison **********\n\n")
+    cat("********** Model Comparison **********\n")
     cat("Absolute PCC difference: ", round(x$pcc_diff, 2), "\n")
     cat("Minimum random PCC difference: ", round(x$min_pcc_difference, digits), "\n")
     cat("Maximum random PCC difference: ", round(x$max_pcc_difference, digits), "\n")
@@ -98,9 +95,6 @@ print.clubprocomparison <- function(x, ...) {
 }
 
 
-# plot.clubprocomparison
-
-
 #' Plot model comparison.
 #'
 #' @details

R/predict.R

@@ -56,18 +56,6 @@ plot.clubpropredictions <- function(x, ...) {
 }
 
 
-#' Print predictions.
-#'
-#' @details
-#' Print a table of observed versus predicted categories
-#'
-#' @param x an object of class "clubpropredictions"
-#' @param ... ignored
-#' @return called for side-effects only
-#' @examples
-#' mod <- club(rate ~ dose, data = caffeine)
-#' z <- predict(mod)
-#' print(z)
 #' @export
 print.clubpropredictions <- function(x, ...) {
     print(unclass(x))

src/cpp_functions.cpp

@@ -18,7 +18,6 @@
 #include <RcppArmadilloExtensions/sample.h>
 #include <progress.hpp>
 #include <progress_bar.hpp>
-using namespace arma;
 // [[Rcpp::depends(RcppArmadillo)]]
 // [[Rcpp::depends(RcppProgress)]]
 
@@ -38,7 +37,7 @@ arma::mat to_indicator_matrix(arma::vec v) {
 // [[Rcpp::export]]
 arma::mat normalise_columns(arma::mat A) {
     for (size_t j {0}; j < A.n_cols; j++) {
-        auto colfactor = sqrt(accu(square(A.col(j))));
+        auto colfactor = std::sqrt(arma::accu(arma::square(A.col(j))));
         for (size_t i {}; i < A.n_rows; i++) {
             A(i, j) = colfactor == 0.0 ? 0.0 : A(i, j) / colfactor;
         }
@@ -50,7 +49,7 @@ arma::mat normalise_columns(arma::mat A) {
 // [[Rcpp::export]]
 arma::mat normalise_rows(arma::mat A) {
     for (size_t i {0}; i < A.n_rows; i++) {
-        auto rowfactor = sqrt(accu(square(A.row(i))));
+        auto rowfactor = std::sqrt(arma::accu(arma::square(A.row(i))));
         for (size_t j {0}; j < A.n_cols; j++) {
             A(i, j) = rowfactor == 0.0 ? 0.0 : A(i, j) / rowfactor;
         }
@@ -62,7 +61,7 @@ arma::mat normalise_rows(arma::mat A) {
 // [[Rcpp::export]]
 arma::mat dichotemise_matrix(arma::mat A) {
     for (size_t i {0}; i < A.n_rows; i++) {
-        auto m = max(A.row(i));
+        auto m = arma::max(A.row(i));
         for(size_t j {0}; j < A.n_cols; j++) {
             A(i, j) = ((A(i, j) == m) && (m > 0)) ? 1.0 : 0.0;
         }
@@ -74,7 +73,7 @@ arma::mat dichotemise_matrix(arma::mat A) {
 // [[Rcpp::export]]
 arma::mat binary_procrustes_rotation(arma::vec obs, arma::mat target_mat, bool normalise_cols) {
     arma::mat obs_mat = to_indicator_matrix(obs);
-    arma::mat T = trans(obs_mat) * target_mat;
+    arma::mat T = arma::trans(obs_mat) * target_mat;
     arma::mat A = normalise_cols ? obs_mat * normalise_rows(normalise_columns(T)) : obs_mat * normalise_rows(T);
     return A;
 }
@@ -95,7 +94,7 @@ double c_pcc(arma::vec obs, arma::mat target_indicator_mat, int imprecision, boo
     arma::mat binary_matrix = dichotemise_matrix(binary_procrustes_rotation(obs, target_indicator_mat, normalise_cols));
     double matches{0.0};
     for (size_t i {0}; i < binary_matrix.n_rows; i++) {
-        if (accu(binary_matrix.row(i)) == 1.0) {
+        if (arma::accu(binary_matrix.row(i)) == 1.0) {
             if (std::abs(int(binary_matrix.row(i).index_max()) - int(target_indicator_mat.row(i).index_max())) <= imprecision) {
                 matches += 1.0;
             }