Package {gbcrosswalk}

Title: Crosswalk Chinese GB Industry Classifications Across Years
Version: 0.0.5
Description: Tools for cleaning, building, and composing crosswalks between Chinese GB/T 4754 industry classification vintages. The package starts with the historical workflow for 1986, 1994, 2002, 2011, and 2017 and is designed so later PDF-derived vintages can be added as additional adjacent pairwise crosswalks.
License: MIT + file LICENSE
URL: https://github.com/robin621/gbcrosswalk
BugReports: https://github.com/robin621/gbcrosswalk/issues
Encoding: UTF-8
RoxygenNote: 7.3.2
Depends: R (≥ 4.1)
Imports: utils
Enhances: readxl
NeedsCompilation: no
Packaged: 2026-07-04 18:28:20 UTC; CodexSandboxOffline
Author: Yajie Wang [aut, cre] (affiliation: Assistant Professor, Division of Global and Area Studies, Chinese University of Hong Kong (Shenzhen))
Maintainer: Yajie Wang <wangyajie621@gmail.com>
Repository: CRAN
Date/Publication: 2026-07-11 08:50:02 UTC

Crosswalk Chinese GB Industry Classifications Across Years

Description

Utilities for cleaning, building, and composing Chinese GB/T 4754 industry classification crosswalks. The package standardizes adjacent pairwise crosswalks into a long format and composes them across years.

Usage

normalize_gb_code(x, width = 4)
normalize_gb_for_level(x, level = c("S", "M", "L"))
codes_to_gb_level(code, level = c("S", "M", "L"))
derive_gb_levels(data, year, s_col = NULL)
fill_down_missing(x)
dominant_prefix(codes, digits = 3, threshold = 0.5,
                special_cutoff = NULL, special_suffix = "")
add_collapsed_targets(data, from_col, to_col, to_all_col = NULL,
                      parent_col = NULL, parent_digits = 3,
                      threshold = 0.5, special_cutoff = NULL,
                      special_suffix = "")
as_gb_pair(data, from_year, to_year, from_col = NULL, to_col = NULL,
           levels = c("S", "M", "L"), source_pair = NULL)
compose_gb_crosswalk(pairs, from_year, to_year, level = c("S", "M", "L"),
                     years = c(1986, 1994, 2002, 2011, 2017))
crosswalk_codes(codes, crosswalk, unmatched = NA_character_)
convert_gb_codes(codes, from_year = NULL, to_year = NULL, level = NULL,
                 pairs = load_gb_crosswalks(),
                 years = c(1986, 1994, 2002, 2011, 2017),
                 unmatched = NA_character_)
detect_gb_year(codes, pairs = load_gb_crosswalks(),
               years = c(1986, 1994, 2002, 2011, 2017),
               level = c("auto", "S", "M", "L"),
               ties = c("all", "first"), details = FALSE)
convert_gb_column(data, column, pairs = load_gb_crosswalks(), from_year, to_year,
                  level = c("S", "M", "L"), output_col = NULL,
                  years = c(1986, 1994, 2002, 2011, 2017),
                  unmatched = NA_character_)
standardize_gb_pairs(pairs)
write_gb_crosswalk_csvs(pairs, dir, prefix = "gb", include_combined = TRUE)
read_gb_crosswalk_csvs(path)
gb_crosswalk_path(file = "gb_all_pairs.csv")
gb_raw_crosswalk_path(file = NULL)
load_gb_crosswalks(path = gb_crosswalk_path())
build_gb_1986_1994(path, sheet = 1)
build_gb_1994_2002(path)
build_gb_2002_2011(path, sheet = 1)
build_gb_2011_2017(path, sheet = 1, old_col = NULL, new_col = NULL)

Arguments

x, code, codes

Character or numeric GB codes.

width

Target code width for left-zero padding.

level

One of "S" for 4-digit, "M" for 3-digit, "L" for 2-digit codes, "auto" where supported, or NULL in convert_gb_codes() to infer it from the input vector.

ties

Whether to return all tied best matches or just the first one.

details

Whether to return all candidate-year scores as well as the best match.

data

A data frame containing crosswalk columns or a user data frame.

column

Name of the input code column to convert.

year, from_year, to_year

GB classification years. In convert_gb_codes(), to_year is required and from_year can be NULL to infer it from the input vector.

s_col, from_col, to_col, old_col, new_col

Column names for source and target codes.

to_all_col, parent_col

Names for collapsed target and dominant parent columns.

parent_digits, digits

Number of prefix digits used for parent codes.

threshold

Share required for a dominant parent prefix.

special_cutoff, special_suffix

Optional fallback rule for service-sector-style parent codes.

levels

Levels to include when standardizing a pairwise crosswalk.

source_pair

Optional source label.

pairs

A standard pair table or list of pair tables. Defaults to the package-shipped merged CSV in convert_gb_column().

file

Package-shipped filename under extdata or raw-crosswalks.

dir

Directory where CSV files should be written.

prefix

Filename prefix for exported CSV files.

include_combined

Whether to also write one combined all-pairs CSV.

output_col

Name of the appended converted-code column.

years

Ordered vector of supported years.

crosswalk

A composed crosswalk with from_code and to_code.

unmatched

Value returned for unmatched input codes.

path

Path to a CSV, XLS, or XLSX crosswalk file, or a directory of CSV files.

sheet

Excel sheet index or name.

Value

Either a character vector, a cleaned data frame, a standard pairwise crosswalk, or a composed crosswalk depending on the function.

Examples

normalize_gb_code(c("123", "0123"))
codes_to_gb_level("0123", "M")
convert_gb_codes(c("572", "593", "843"), from_year = 2017, to_year = 2011, level = "M")
convert_gb_codes(c("572", "593", "843"), to_year = 2011)