Prepare for cran

R/ext.R

@@ -15,6 +15,7 @@ a <- b <- NULL
 #'
 #' @return named list with entries `"a"`, the location parameter, and `"b"`, the
 #'   scale parameter
+#' @noRd
 calc_loc_scale_params <- function(ps, qs, dist) {
   if (dist == "lnorm") {
     if (any(qs <= 0.0)) {
@@ -48,6 +49,7 @@ calc_loc_scale_params <- function(ps, qs, dist) {
 #'   evaluate the density function (or its log) of the distribution in the
 #'   specified location-scale family that has quantiles matching those in `ps`
 #'   and `qs`
+#' @noRd
 d_ext_factory <- function(ps, qs, dist) {
   c(a, b) %<-% calc_loc_scale_params(ps, qs, dist)
 
@@ -88,6 +90,7 @@ d_ext_factory <- function(ps, qs, dist) {
 #'   evaluate the cumulative distribution function (or its log) of the
 #'   distribution in the specified location-scale family that has quantiles
 #'   matching those in `ps` and `qs`
+#' @noRd
 p_ext_factory <- function(ps, qs, dist) {
   c(a, b) %<-% calc_loc_scale_params(ps, qs, dist)
 
@@ -123,6 +126,7 @@ p_ext_factory <- function(ps, qs, dist) {
 #' @return a function with parameter `p` that can be used to evaluate the
 #'   quantile function of the distribution in the specified location-scale
 #'   family that has quantiles matching those in `ps` and `qs`
+#' @noRd
 q_ext_factory <- function(ps, qs, dist) {
   c(a, b) %<-% calc_loc_scale_params(ps, qs, dist)
 

R/int.R

@@ -2,13 +2,11 @@
 #' interpolating a given set of points.
 #'
 #' @param x vector giving the x coordinates of the points to be interpolated.
-#'   Alternatively a single plotting structure can be specified: see ‘xy.coords’.
 #'
 #' @param y vector giving the y coordinates of the points to be interpolated.
-#'   Must be increasing or decreasing for ‘method = "hyman"’. Alternatively a
-#'   single plotting structure can be specified: see ‘xy.coords’.
+#'   Must be increasing or decreasing for 'method = "hyman"'.
 #'
-#' @param m (for ‘splinefunH()’) vector of _slopes_ \eqn{m_i}{m[i]} at the
+#' @param m (for 'splinefunH()') vector of _slopes_ \eqn{m_i}{m[i]} at the
 #'   points \eqn{(x_i,y_i)}{(x[i],y[i])}; these together determine the
 #'   *H*ermite “spline” which is piecewise cubic, (only) _once_ differentiable
 #'   continuously.
@@ -103,7 +101,7 @@ mono_Hermite_spline <- function(x, y, m) {
 #' @param y numeric vector with the "vertical axis" coordinates of the points
 #'   to interpolate.
 #' @param cont_dir at steps or discontinuities, the direction from which the
-#'   function is continuous. This will be "right" for a cdf or "left" for a qf.
+#'   function is continuous. This will be "right" for a CDF or "left" for a QF.
 #' @param increasing boolean indicating whether the function is increasing or
 #'   decreasing. Only used in the degenerate case where there is only one unique
 #'   value of `x`.
@@ -202,41 +200,41 @@ step_interp_factory <- function(x, y, cont_dir = c("right", "left"),
 }
 
 
-#' Approximate density function, cdf, or quantile function on the interior of
+#' Approximate density function, CDF, or quantile function on the interior of
 #' provided quantiles by representing the distribution as a sum of a discrete
-#' part at any duplicated `qs` and a continuous part for which the cdf is
+#' part at any duplicated `qs` and a continuous part for which the CDF is
 #' estimated using a monotonic Hermite spline. See details for more.
 #'
 #' @param ps vector of probability levels
-#' @param qs vector of quantile values correponding to ps
+#' @param qs vector of quantile values corresponding to ps
 #' @param tail_dist name of parametric distribution for the tails
-#' @param fn_type the type of function that is requested: `"d"` for a pdf,
-#'   `"p"` for a cdf, or `"q"` for a quantile function.
+#' @param fn_type the type of function that is requested: `"d"` for a PDF,
+#'   `"p"` for a CDF, or `"q"` for a quantile function.
 #' @param n_grid grid size to use when augmenting the input `qs` to obtain a
 #'   finer grid of points along which we form a piecewise linear approximation
 #'   to the spline. `n_grid` evenly-spaced points are inserted between each
 #'   pair of consecutive values in `qs`. The default value is 20. This can
 #'   be set to `NULL`, in which case the piecewise linear approximation is not
 #'   used. This is not recommended if the `fn_type` is `"q"`.
 #'
-#' @details The cdf of the continuous part of the distribution is estimated
+#' @details The CDF of the continuous part of the distribution is estimated
 #' using a monotonic degree 3 Hermite spline that interpolates the quantiles
 #' after subtracting the discrete distribution and renormalizing. In theory,
 #' an estimate of the quantile function could be obtained by directly inverting
 #' this spline. However, in practice, we have observed that this can suffer from
 #' numerical problems. Therefore, the default behavior of this function is to
-#' evaluate the "stage 1" cdf estimate corresponding to discrete point masses
-#' plus monotonic spline at a fine grid of points, and use the "stage 2" cdf
+#' evaluate the "stage 1" CDF estimate corresponding to discrete point masses
+#' plus monotonic spline at a fine grid of points, and use the "stage 2" CDF
 #' estimate that linearly interpolates these points with steps at any duplicated
 #' q values. The quantile function estimate is obtained by inverting this
-#' "stage 2" cdf estimate. When the distribution is continuous, we can obtain an
-#' estimate of the pdf by differentiating the cdf estimate, resulting in a
+#' "stage 2" CDF estimate. When the distribution is continuous, we can obtain an
+#' estimate of the PDF by differentiating the CDF estimate, resulting in a
 #' discontinuous "histogram density". The size of the grid can be specified with
 #' the `n_grid` argument. In settings where it is desirable to obtain a
-#' continuous density function, the "stage 1" cdf estimate can be used by
+#' continuous density function, the "stage 1" CDF estimate can be used by
 #' setting `n_grid = NULL`.
 #'
-#' @return a function to evaluate the pdf, cdf, or quantile function.
+#' @return a function to evaluate the PDF, CDF, or quantile function.
 #' @rdname spline_cdf
 spline_cdf <- function(ps, qs, tail_dist,
                        fn_type = c("d", "p", "q"),
@@ -250,7 +248,7 @@ spline_cdf <- function(ps, qs, tail_dist,
   }
 }
 
-#' @rdname spline_cdf
+#' @noRd
 #' @importFrom stats approx
 spline_cdf_grid_interp <- function(ps, qs, tail_dist,
                                    fn_type = c("d", "p", "q"),
@@ -291,6 +289,7 @@ spline_cdf_grid_interp <- function(ps, qs, tail_dist,
 #' pair of consecutive values in qs, and the corresponding ps are filled in
 #' by evaluating the spline output from `spline_cdf_direct` at the qs grid.
 #' @keywords internal
+#' @noRd
 grid_augment_ps_qs <- function(ps, qs, tail_dist, n_grid) {
   n_grid <- as.integer(n_grid)
   if (n_grid < 0) {
@@ -328,14 +327,14 @@ grid_augment_ps_qs <- function(ps, qs, tail_dist, n_grid) {
 #' Internal function that constructs a monotonic Hermite spline interpolating
 #' ps and qs.
 #' @importFrom stats predict
-#' @rdname spline_cdf
+#' @noRd
 spline_cdf_direct <- function(ps, qs, tail_dist,
                               fn_type = c("d", "p", "q")) {
   # fit a monotonic spline to the qs and ps for the continuous part of the
   # distribution to approximate the cdf on the interior
   # the vector m that we assemble here has the target slopes of the spline
   # at each data point (qs[i], ps[i])
-  # on ends, slope of cdf approximation should match tail distribution pdf
+  # on ends, slope of cdf approximation should match tail distribution PDF
   # on interior, slope is the mean of the slopes of the adjacent segments
   # note: there is probably room for improvement for this interior behavior,
   # but this seems to be the standard strategy for monotonic splines

R/make_dist_fns.R

@@ -9,9 +9,10 @@ disc_weight <- disc_ps <- disc_qs <- cont_ps <- cont_qs <- disc_ps_range <- NULL
 #' Clean up ps and qs provided by user: handle missing and unsorted values
 #'
 #' @param ps vector of probability levels
-#' @param qs vector of quantile values correponding to ps
+#' @param qs vector of quantile values corresponding to ps
 #'
 #' @return named list with entries `ps` and `qs`
+#' @noRd
 clean_ps_and_qs <- function(ps, qs) {
   checkmate::assert_numeric(ps, lower = 0, upper = 1)
   checkmate::assert_numeric(qs)
@@ -40,7 +41,7 @@ clean_ps_and_qs <- function(ps, qs) {
 #' from a set of quantiles of the distribution.
 #'
 #' @param ps vector of probability levels
-#' @param qs vector of quantile values correponding to ps
+#' @param qs vector of quantile values corresponding to ps
 #' @param interior_method method for interpolating the distribution on the
 #'   interior of the provided `qs`. This package provides one method for this,
 #'   `"spline_cdf"`. The user may also provide a custom function; see the
@@ -60,8 +61,8 @@ clean_ps_and_qs <- function(ps, qs) {
 #'   distribution as a sum of a discrete component at any points where there
 #'   are duplicated `qs` for multiple different `ps` and a continuous component
 #'   that is estimated by using a monotonic cubic spline that interpolates the
-#'   provided `(q, p)` pairs as an estimate of the cdf. The density function is
-#'   then obtained by differentiating this estimate of the cdf.
+#'   provided `(q, p)` pairs as an estimate of the CDF. The density function is
+#'   then obtained by differentiating this estimate of the CDF.
 #'
 #'   Optionally, the user may provide another function that accepts arguments
 #'   `ps`, `qs`, `tail_dist`, and `fn_type` (which will be either `"d"`, `"p"`,
@@ -149,7 +150,7 @@ make_d_fn <- function(ps, qs,
 #' from a set of quantiles of the distribution.
 #'
 #' @param ps vector of probability levels
-#' @param qs vector of quantile values correponding to ps
+#' @param qs vector of quantile values corresponding to ps
 #' @param interior_method method for interpolating the distribution on the
 #'   interior of the provided `qs`. This package provides one method for this,
 #'   `"spline_cdf"`. The user may also provide a custom function; see the
@@ -169,7 +170,7 @@ make_d_fn <- function(ps, qs,
 #'   distribution as a sum of a discrete component at any points where there
 #'   are duplicated `qs` for multiple different `ps` and a continuous component
 #'   that is estimated by using a monotonic cubic spline that interpolates the
-#'   provided `(q, p)` pairs as an estimate of the cdf.
+#'   provided `(q, p)` pairs as an estimate of the CDF.
 #'
 #'   Optionally, the user may provide another function that accepts arguments
 #'   `ps`, `qs`, `tail_dist`, and `fn_type` (which will be either `"d"`, `"p"`,
@@ -294,7 +295,7 @@ make_p_fn <- function(ps, qs,
 #' quantiles of the distribution.
 #'
 #' @param ps vector of probability levels
-#' @param qs vector of quantile values correponding to ps
+#' @param qs vector of quantile values corresponding to ps
 #' @param interior_method method for interpolating the distribution on the
 #'   interior of the provided `qs`. This package provides one method for this,
 #'   `"spline_cdf"`. The user may also provide a custom function; see the
@@ -314,8 +315,8 @@ make_p_fn <- function(ps, qs,
 #'   distribution as a sum of a discrete component at any points where there
 #'   are duplicated `qs` for multiple different `ps` and a continuous component
 #'   that is estimated by using a monotonic cubic spline that interpolates the
-#'   provided `(q, p)` pairs as an estimate of the cdf. The quantile function
-#'   is then obtained by inverting this estimate of the cdf.
+#'   provided `(q, p)` pairs as an estimate of the CDF. The quantile function
+#'   is then obtained by inverting this estimate of the CDF.
 #'
 #'   Optionally, the user may provide another function that accepts arguments
 #'   `ps`, `qs`, `tail_dist`, and `fn_type` (which will be either `"d"`, `"p"`,
@@ -443,7 +444,7 @@ make_q_fn <- function(ps, qs,
 #' quantiles of the distribution.
 #'
 #' @param ps vector of probability levels
-#' @param qs vector of quantile values correponding to ps
+#' @param qs vector of quantile values corresponding to ps
 #' @param interior_method method for interpolating the distribution on the
 #'   interior of the provided `qs`. This package provides one method for this,
 #'   `"spline_cdf"`. The user may also provide a custom function; see the
@@ -463,8 +464,8 @@ make_q_fn <- function(ps, qs,
 #'   distribution as a sum of a discrete component at any points where there
 #'   are duplicated `qs` for multiple different `ps` and a continuous component
 #'   that is estimated by using a monotonic cubic spline that interpolates the
-#'   provided `(q, p)` pairs as an estimate of the cdf. The quantile function
-#'   is then obtained by inverting this estimate of the cdf.
+#'   provided `(q, p)` pairs as an estimate of the CDF. The quantile function
+#'   is then obtained by inverting this estimate of the CDF.
 #'
 #'   Optionally, the user may provide another function that accepts arguments
 #'   `ps`, `qs`, `tail_dist`, and `fn_type` (which will be either `"d"`, `"p"`,

R/utils.R

@@ -80,7 +80,7 @@
 #' parts of a distribution.
 #'
 #' @param ps vector of probability levels
-#' @param qs vector of quantile values correponding to ps
+#' @param qs vector of quantile values corresponding to ps
 #' @param dup_tol numeric tolerance for identifying duplicated values indicating
 #'   a discrete component of the distribution. If there is a run of values where
 #'   each consecutive pair is closer together than the tolerance, all are
@@ -103,15 +103,15 @@
 #'   - `disc_qs`: locations of discrete components, corresponding to duplicated
 #'     values in the input `qs`. May be `numeric(0)` if there are no discrete
 #'     components.
-#'   - `cont_ps`: probability levels for the continous part of the distribution
-#'   - `cont_qs`: quantile values for the continous part of the distribution
+#'   - `cont_ps`: probability levels for the continuous part of the distribution
+#'   - `cont_qs`: quantile values for the continuous part of the distribution
 #'   - `disc_ps_range`: a list of length equal to the number of point masses in
 #'     the discrete distribution. Each entry is a numeric vector of length two
-#'     with the value of the cdf approaching the point mass from the left and
+#'     with the value of the CDF approaching the point mass from the left and
 #'     from the right.
 #'
 #' @export
 split_disc_cont_ps_qs <- function(ps, qs, dup_tol = 1e-6, zero_tol = 1e-12,
                                   is_hurdle = FALSE) {
   # if zero counts as discrete and any qs are (approximately) zero, augment
   # with an additional zero so that logic below based on duplicate qs works