New factor functions.

NAMESPACE

@@ -29,6 +29,7 @@ S3method(sset,sf)
 S3method(sset,tbl_df)
 export("%!in_%")
 export("%in_%")
+export(add_na_level)
 export(all_na)
 export(any_na)
 export(col_all_na)
@@ -37,6 +38,8 @@ export(col_na_counts)
 export(count_val)
 export(cut_numeric)
 export(deframe_)
+export(drop_levels)
+export(drop_na_level)
 export(enframe_)
 export(factor_)
 export(gcd)

R/factors.R

@@ -1,11 +1,15 @@
 #' A faster version of `factor()`
 #'
 #' @description
-#' A fast version of `factor()` using the collapse package.
+#' A fast version of `factor()` using the collapse package. \cr
+#'
 #' There are some additional utilities such as
 #' `levels_factor()` which returns the levels of a factor, as a factor,
 #' `used_levels()` which returns the used levels of a factor,
-#' and `unused_levels()` which returns the unused levels of a factor.
+#' `unused_levels()` which returns the unused levels of a factor,
+#' `add_na_level()` which adds an explicit `NA` level,
+#' `drop_na_level()` which drops the `NA` level,
+#' and `drop_levels()` which drops unused factor levels.
 #'
 #' @returns
 #' A `factor` or `character` in the case of `used_levels` and `unused_levels`.
@@ -18,6 +22,8 @@
 #' @param na_exclude Should `NA` values be excluded from the factor levels?
 #' Default is `TRUE`.
 #' @param ordered Should the result be an ordered factor?
+#' @param name Name of `NA` level.
+#' @param where Where should `NA` level be placed? Either first or last.
 #'
 #' @details
 #' This operates similarly to `collapse::qF()`. \cr
@@ -69,12 +75,18 @@ factor_ <- function(x = integer(), levels = NULL, order = TRUE,
 #' @export
 #' @rdname factors
 levels_factor <- function(x){
+  check_is_factor(x)
   lvls <- levels(x)
   out <- seq_along(lvls)
   attr(out, "levels") <- lvls
   class(out) <- class(x)
   out
 }
+check_is_factor <- function(x){
+  if (!is.factor(x)){
+    stop("x must be a factor")
+  }
+}
 #' @export
 #' @rdname factors
 used_levels <- function(x){
@@ -85,3 +97,66 @@ used_levels <- function(x){
 unused_levels <- function(x){
   as.character(setdiff_(levels_factor(x), x))
 }
+#' @export
+#' @rdname factors
+add_na_level <- function(x, name = NA, where = c("last", "first")){
+  check_is_factor(x)
+  where <- match.arg(where)
+  lvls <- levels(x)
+  if (any_na(lvls)){
+    x
+  } else {
+    out <- unclass(x)
+    n_lvls <- length(lvls)
+
+    if (where == "first"){
+      out <- out + 1L
+      attr(out, "levels") <- c(name, lvls)
+      out[which_na(out)] <- 1L
+    } else {
+      attr(out, "levels") <- c(lvls, name)
+      out[which_na(out)] <- n_lvls + 1L
+    }
+
+    class(out) <- class(x)
+    out
+  }
+}
+#' @export
+#' @rdname factors
+drop_na_level <- function(x){
+
+  check_is_factor(x)
+  lvls <- levels(x)
+
+  which_na_lvl <- which_na(lvls)
+  if (length(which_na_lvl) == 0){
+    x
+  } else {
+    new_lvls <- lvls[-which_na_lvl]
+
+    matches <- collapse::fmatch(lvls, new_lvls, overid = 2L)
+    out <- matches[unclass(x)]
+
+    attributes(out) <- attributes(x)
+    attr(out, "levels") <- new_lvls
+
+    out
+  }
+}
+#' @export
+#' @rdname factors
+drop_levels <- function(x){
+  lvls <- levels(x)
+  n_lvls <- length(lvls)
+  used_lvls <- intersect_(levels_factor(x), x)
+  if (length(used_lvls) == n_lvls){
+    x
+  } else {
+   factor_(x, levels = used_lvls)
+    # Alternative
+    # out <- collapse::fmatch(lvls, used_lvls, overid = 2L)[unclass(x)]
+    # attributes(out) <- attributes(x)
+    # attr(out, "levels") <- as.character(used_lvls)
+  }
+}

src/nas.cpp

@@ -1,57 +1,5 @@
 #include "cheapr_cpp.h"
 
-// The below works but probably not a good idea to use internal R code
-// Direct inclusion of internal structure definitions
-// typedef struct sxpinfo_struct {
-//   SEXPTYPE type      : 5;
-//   unsigned int obj   : 1;
-//   unsigned int named : 2;
-//   unsigned int gp    : 16;
-//   unsigned int mark  : 1;
-//   unsigned int debug : 1;
-//   unsigned int trace : 1;
-//   unsigned int spare : 1;
-//   unsigned int gcgen : 1;
-//   unsigned int gccls : 3;
-// } sxpinfo_struct;
-//
-// typedef struct SEXPREC {
-//   sxpinfo_struct sxpinfo;
-//   struct SEXPREC *attrib;
-//   struct SEXPREC *gengc_next_node, *gengc_prev_node;
-//   union {
-//     struct primsxp_struct {int offset; } primsxp;
-//     struct symsxp_struct {
-//       struct SEXPREC *pname;
-//       struct SEXPREC *value;
-//       struct SEXPREC *internal;
-//     } symsxp;
-//     struct listsxp_struct {
-//       struct SEXPREC *carval;
-//       struct SEXPREC *cdrval;
-//       struct SEXPREC *tagval;
-//     } listsxp;
-//     struct envsxp_struct {
-//       struct SEXPREC *frame;
-//       struct SEXPREC *enclos;
-//       struct SEXPREC *hashtab;
-//     } envsxp;
-//     struct closxp_struct {
-//       struct SEXPREC *formals;
-//       struct SEXPREC *body;
-//       struct SEXPREC *env;
-//     } closxp;
-//     struct promsxp_struct {
-//       struct SEXPREC *value;
-//       struct SEXPREC *expr;
-//       struct SEXPREC *env;
-//     } promsxp;
-//   } u;
-// } SEXPREC, *SEXP;
-//
-// #define SET_TYPEOF2(x, v) ((x)->sxpinfo.type = (v))
-
-
 R_xlen_t na_count(SEXP x, bool recursive){
   R_xlen_t n = Rf_xlength(x);
   R_xlen_t count = 0;