Skip to contents

Tidy the cell-status (missing-cell) table

Source: R/read.R
ivt_tidy_missing.Rd

The companion of ivt_tidy() for x$missing: labels the coordinate columns exactly as ivt_tidy() labels x$cells, so the two tables line up column for column and the missing cells can be joined onto – or unioned with – the values. Requires read_ivt(path, missing = TRUE).

Usage

ivt_tidy_missing(
  x,
  labels = TRUE,
  trim_labels = TRUE,
  dim_names = c("slug", "label"),
  language = "en",
  depth = FALSE
)

Arguments

x

An ivt object from read_ivt().

labels

If TRUE (default) replace member-id columns with member labels; if FALSE return the compact integer-id table (member ids).

trim_labels

If TRUE (default) strip the hierarchy-indentation spaces from member labels.

dim_names

How to name the data-dimension columns: "slug" (default) uses the terse structural slug (e.g. age), which is compact and language-neutral; "label" uses the full dimension name (e.g. Age of primary household maintainer, or its French equivalent when language = "fr"). Slug output can be labelled afterwards with label_ivt_columns(). The choice applies to both labels values.

language

Output language for labels and label-derived column names: "en" (default) or "fr". Also accepts "eng"/"fra" and any case (it is lower-cased). French falls back to English wherever the file carries no French copy (e.g. the language-neutral geo_uid, or a dimension with no French name).

depth

If TRUE (default FALSE) add a <col>_depth integer column after each data-dimension column giving that member's hierarchy depth (read from the label indentation, the same measure carried by ivt_members()). Opt-in, so the default output – and hence the Parquet written by ivt_write_parquet() – is unchanged.

Value

A tibble: the coordinate columns of ivt_tidy(), then symbol and status.

Details

The result has no value column (these cells have no value); in its place are symbol and status, the reason the file states for the absence, or NA where the page carries only the bare absent mask (missing, cause not stated). The file's own reason-code legend rides along as attr(., "legend"), and the per-page-class tally as attr(., "pages").

See also

read_ivt() (the "Missing values" section), ivt_write_parquet(), which writes this table as a _missing.parquet sidecar.

Examples

path <- system.file("extdata", "98100044.ivt", package = "canivt")
ivt <- read_ivt(path, missing = TRUE)
ivt_tidy_missing(ivt)
#> # A tibble: 0 × 8
#> # ℹ 8 variables: geo_label <chr>, geo_name <chr>, geo_uid <chr>,
#> #   geo_level <chr>, type <chr>, collective <chr>, symbol <chr>, status <chr>