Updated sset and news.

NEWS.md

@@ -1,5 +1,8 @@
 # cheapr (Development version)
 
+* New function `sset` to consistently subset data frame rows and vectors in 
+general.
+
 * `overview` now always returns an object of class "overview". It also returns
 the number of observations instead of rows so that it makes sense 
 for vector summaries as well as data frame summaries.

R/cpp11.R

@@ -128,6 +128,10 @@ cpp_list_as_df <- function(x) {
   .Call(`_cheapr_cpp_list_as_df`, x)
 }
 
+cpp_set_rm_attributes <- function(x) {
+  .Call(`_cheapr_cpp_set_rm_attributes`, x)
+}
+
 cpp_set_copy_attributes <- function(target, source, attrs) {
   .Call(`_cheapr_cpp_set_copy_attributes`, target, source, attrs)
 }

R/sset.R

@@ -7,7 +7,6 @@
 #'
 #' @param x Vector or data frame.
 #' @param i A logical or vector of indices. \cr
-#' The default is 0 which differs to `[`.
 #' @param j Column indices, names or logical vector.
 #' @param ... Further parameters passed to `[`.
 #'
@@ -19,6 +18,19 @@
 #' @examples
 #' library(cheapr)
 #' library(bench)
+#'
+#' # Selecting columns
+#' sset(airquality, j = "Temp")
+#' sset(airquality, j = 1:2)
+#'
+#' # Selecting rows
+#' sset(iris, 1:5)
+#'
+#' # Rows and columns
+#' sset(iris, 1:5, 1:5)
+#' sset(iris, iris$Sepal.Length > 7, c("Species", "Sepal.Length"))
+#'
+#' # Cnmparison against base
 #' x <- rnorm(10^4)
 #'
 #' mark(x[1:10^3], sset(x, 1:10^3))
@@ -32,47 +44,53 @@
 #'
 #' @rdname sset
 #' @export
-sset <- function(x, i = 0, ...){
+sset <- function(x, ...){
   UseMethod("sset")
 }
 #' @export
-sset.default <- function(x, i = 0, ...){
-  if (is.logical(i)){
+sset.default <- function(x, i, ...){
+  if (!missing(i) && is.logical(i)){
     check_length(i, length(x))
     i <- which_(i)
   }
   x[i, ...]
 }
 #' @rdname sset
 #' @export
-sset.data.frame <- function(x, i = 0, j = seq_along(x), ...){
+sset.data.frame <- function(x, i, j = seq_along(x), ...){
   df_subset(x, i, j)
 }
 #' @rdname sset
 #' @export
-sset.tbl_df <- function(x, i = 0, j = seq_along(x), ...){
+sset.tbl_df <- function(x, i, j = seq_along(x), ...){
   out <- df_subset(x, i, j)
   class(out) <- c("tbl_df", "tbl", "data.frame")
   out
 }
 #' @rdname sset
 #' @export
-sset.POSIXlt <- function(x, i = 0, ...){
+sset.POSIXlt <- function(x, i, ...){
   out <- df_subset(list_as_df(x), i)
-  cpp_set_copy_attributes(out, x, names(attributes(x)))
+  cpp_set_copy_attributes(
+    cpp_set_rm_attributes(out), x, names(attributes(x))
+  )
 }
 #' @rdname sset
 #' @export
-sset.data.table <- function(x, i = 0, j = seq_along(x), ...){
-  # collapse::qDT(df_subset(x, i, j))
+sset.data.table <- function(x, i, j = seq_along(x), ...){
+  # This is to ensure that a copy is made basically
+  # More efficient to use data.table::copy()
+  if (missing(i)){
+    i <- seq_len(nrow(x))
+  }
   out <- df_subset(x, i, j)
   cpp_set_copy_attributes(
     out, x, c("class", ".internal.selfref")
   )
 }
 #' @rdname sset
 #' @export
-sset.sf <- function(x, i = 0, j = seq_along(x), ...){
+sset.sf <- function(x, i, j = seq_along(x), ...){
   out <- df_subset(x, i, j)
   source_nms <- names(attributes(x))
   invisible(
@@ -81,3 +99,44 @@ sset.sf <- function(x, i = 0, j = seq_along(x), ...){
   class(out) <- class(x)
   out
 }
+df_select <- function(x, i){
+  attrs <- attributes(x)
+  out <- cpp_list_rm_null(unclass(x)[i])
+  attrs[["names"]] <- attr(out, "names")
+  attrs[["row.names"]] <- .row_names_info(x, type = 0L)
+  attributes(out) <- attrs
+  out
+}
+
+# Efficient data frame subset
+# With the exception of which_() this is surprisingly all base R...
+# It relies on sset which falls back on `[` when no method is found.
+df_subset <- function(x, i, j = seq_along(x)){
+  missingi <- missing(i)
+  nrows <- length(attr(x, "row.names"))
+  if (!missingi && is.logical(i)){
+    check_length(i, nrows)
+    i <- which_(i)
+  }
+
+  ### Subset columns
+
+  out <- df_select(x, j)
+
+  ### Subset rows
+
+  if (!missingi){
+    if (length(out) == 0){
+      attr(out, "row.names") <- .set_row_names(
+        length(
+          seq_len(nrows)[i]
+        )
+      )
+    } else {
+      out <- list_as_df(
+        lapply(out, sset, i)
+      )
+    }
+  }
+  out
+}

R/utils.R

@@ -76,14 +76,6 @@ which_in <- function(x, table){
 which_not_in <- function(x, table){
   which_na(collapse::fmatch(x, table, overid = 2L, nomatch = NA_integer_))
 }
-df_select <- function(x, i){
-  attrs <- attributes(x)
-  out <- cpp_list_rm_null(unclass(x)[i])
-  attrs[["names"]] <- attr(out, "names")
-  attrs[["row.names"]] <- .row_names_info(x, type = 0L)
-  attributes(out) <- attrs
-  out
-}
 tzone <- function(x){
   out <- attr(x, "tzone")
   if (is.null(out)) {
@@ -94,33 +86,6 @@ tzone <- function(x){
   }
 }
 
-# Efficient data frame subset
-# With the exception of which_() this is surprisingly all base R...
-# It relies on sset which falls back on `[` when no method is found.
-df_subset <- function(x, i = 0, j = seq_along(x)){
-  nrows <- length(attr(x, "row.names"))
-  if (is.logical(i)){
-    check_length(i, nrows)
-    i <- which_(i)
-  }
-  l <- as.list(
-    df_select(x, j)
-  )
-  if (length(l) == 0){
-    out <- list_as_df(l)
-    attr(out, "row.names") <- .set_row_names(
-      length(
-        seq_len(nrows)[i]
-      )
-    )
-  } else {
-    out <- list_as_df(
-      lapply(l, sset, i)
-    )
-  }
-  out
-}
-
 # safe_unique <- function(x, ...){
 #   out <- tryCatch(collapse::funique(x, ...), error = function(e) return(".r.error"))
 #   if (length(out) == 1 && out == ".r.error"){

src/cpp11.cpp

@@ -230,6 +230,13 @@ extern "C" SEXP _cheapr_cpp_list_as_df(SEXP x) {
   END_CPP11
 }
 // utils.cpp
+SEXP cpp_set_rm_attributes(SEXP x);
+extern "C" SEXP _cheapr_cpp_set_rm_attributes(SEXP x) {
+  BEGIN_CPP11
+    return cpp11::as_sexp(cpp_set_rm_attributes(cpp11::as_cpp<cpp11::decay_t<SEXP>>(x)));
+  END_CPP11
+}
+// utils.cpp
 SEXP cpp_set_copy_attributes(SEXP target, SEXP source, SEXP attrs);
 extern "C" SEXP _cheapr_cpp_set_copy_attributes(SEXP target, SEXP source, SEXP attrs) {
   BEGIN_CPP11
@@ -275,6 +282,7 @@ static const R_CallMethodDef CallEntries[] = {
     {"_cheapr_cpp_row_na_counts",        (DL_FUNC) &_cheapr_cpp_row_na_counts,        1},
     {"_cheapr_cpp_sequence",             (DL_FUNC) &_cheapr_cpp_sequence,             3},
     {"_cheapr_cpp_set_copy_attributes",  (DL_FUNC) &_cheapr_cpp_set_copy_attributes,  3},
+    {"_cheapr_cpp_set_rm_attributes",    (DL_FUNC) &_cheapr_cpp_set_rm_attributes,    1},
     {"_cheapr_cpp_vec_length",           (DL_FUNC) &_cheapr_cpp_vec_length,           1},
     {"_cheapr_cpp_which_",               (DL_FUNC) &_cheapr_cpp_which_,               2},
     {"_cheapr_cpp_which_na",             (DL_FUNC) &_cheapr_cpp_which_na,             1},

src/utils.cpp

@@ -165,6 +165,21 @@ SEXP cpp_list_as_df(SEXP x) {
   }
 }
 
+// Remove attributes in-place
+
+[[cpp11::register]]
+SEXP cpp_set_rm_attributes(SEXP x){
+  SEXP attrs = Rf_protect(cpp11::package("base")["attributes"](x));
+  SEXP names = Rf_protect(Rf_getAttrib(attrs, R_NamesSymbol));
+  int n = Rf_length(attrs);
+  for (int i = 0; i < n; ++i){
+    SEXP attrib_nm = Rf_protect(Rf_install(CHAR(STRING_ELT(names, i))));
+    Rf_setAttrib(x, attrib_nm, R_NilValue);
+  }
+  Rf_unprotect(n + 2);
+  return x;
+}
+
 // Copy specified attributes (character vector of names)
 // from source to target (by reference)
 // Use with extreme care as it modifies target in-place