Help documentation updates
Update slide.R help doc
irkaal
| 1 |
#' Constructor for new rset objects
|
|
| 2 |
#' @param splits A list column of `rsplits` or a tibble with a single column
|
|
| 3 |
#' called "splits" with a list column of `rsplits`.
|
|
| 4 |
#' @param ids A character vector or a tibble with one or more columns that
|
|
| 5 |
#' begin with "id".
|
|
| 6 |
#' @param attrib An optional named list of attributes to add to the object.
|
|
| 7 |
#' @param subclass A character vector of subclasses to add.
|
|
| 8 |
#' @return An `rset` object.
|
|
| 9 |
#' @keywords internal
|
|
| 10 |
#' @export
|
|
| 11 |
new_rset <- function(splits, ids, attrib = NULL, |
|
| 12 |
subclass = character()) { |
|
| 13 | 1 |
stopifnot(is.list(splits)) |
| 14 | 1 |
if (!is_tibble(ids)) { |
| 15 | 1 |
ids <- tibble(id = ids) |
| 16 |
} else { |
|
| 17 | 1 |
if (!all(grepl("^id", names(ids)))) { |
| 18 | 1 |
rlang::abort("The `ids` tibble column names should start with 'id'.") |
| 19 |
}
|
|
| 20 |
}
|
|
| 21 | 1 |
either_type <- function(x) |
| 22 | 1 |
is.character(x) | is.factor(x) |
| 23 | 1 |
ch_check <- vapply(ids, either_type, c(logical = TRUE)) |
| 24 | 1 |
if (!all(ch_check)) { |
| 25 | 1 |
rlang::abort("All ID columns should be character or factor vectors.") |
| 26 |
}
|
|
| 27 |
|
|
| 28 | 1 |
if (!is_tibble(splits)) { |
| 29 | 1 |
splits <- tibble(splits = splits) |
| 30 |
} else { |
|
| 31 |
if (ncol(splits) > 1 | names(splits)[1] != "splits") { |
|
| 32 |
rlang::abort( |
|
| 33 |
"The `splits` tibble should have a single column named `splits`."
|
|
| 34 |
)
|
|
| 35 |
}
|
|
| 36 |
}
|
|
| 37 |
|
|
| 38 | 1 |
if (nrow(ids) != nrow(splits)) { |
| 39 | 1 |
rlang::abort("Split and ID vectors have different lengths.") |
| 40 |
}
|
|
| 41 |
|
|
| 42 |
# Create another element to the splits that is a tibble containing
|
|
| 43 |
# an identifier for each id column so that, in isolation, the resample
|
|
| 44 |
# id can be known just based on the `rsplit` object. This can then be
|
|
| 45 |
# accessed using the `labels` method for `rsplits`
|
|
| 46 |
|
|
| 47 | 1 |
splits$splits <- map2( |
| 48 | 1 |
splits$splits, |
| 49 | 1 |
split_unnamed(ids, rlang::seq2(1L, nrow(ids))), |
| 50 | 1 |
add_id |
| 51 |
)
|
|
| 52 |
|
|
| 53 | 1 |
res <- bind_cols(splits, ids) |
| 54 |
|
|
| 55 | 1 |
if (!is.null(attrib)) { |
| 56 | 1 |
if (any(names(attrib) == "")) { |
| 57 | 1 |
rlang::abort("`attrib` should be a fully named list.") |
| 58 |
}
|
|
| 59 | 1 |
for (i in names(attrib)) { |
| 60 | 1 |
attr(res, i) <- attrib[[i]] |
| 61 |
}
|
|
| 62 |
}
|
|
| 63 |
|
|
| 64 | 1 |
if (length(subclass) > 0) { |
| 65 | 1 |
res <- add_class(res, cls = subclass, at_end = FALSE) |
| 66 |
}
|
|
| 67 |
|
|
| 68 | 1 |
res |
| 69 |
}
|
|
| 70 |
|
|
| 71 |
add_id <- function(split, id) { |
|
| 72 | 1 |
split$id <- id |
| 73 | 1 |
split
|
| 74 |
}
|
|
| 75 |
|
|
| 76 |
# ------------------------------------------------------------------------------
|
|
| 77 |
|
|
| 78 |
#' @export
|
|
| 79 |
`[.rset` <- function(x, i, j, drop = FALSE, ...) { |
|
| 80 | 1 |
out <- NextMethod() |
| 81 | 1 |
rset_reconstruct(out, x) |
| 82 |
}
|
|
| 83 |
|
|
| 84 |
# ------------------------------------------------------------------------------
|
|
| 85 |
|
|
| 86 |
#' @export
|
|
| 87 |
`names<-.rset` <- function(x, value) { |
|
| 88 | 1 |
out <- NextMethod() |
| 89 | 1 |
rset_reconstruct(out, x) |
| 90 |
}
|
Read our documentation on viewing source code .