Package 'klassR'

Description

Match and convert a classification

Usage

apply_klass(
  x,
  classification,
  date = NULL,
  variant = NULL,
  correspond = NULL,
  language = "nb",
  output_level = NULL,
  output = "name",
  format = TRUE
)

ApplyKlass(
  x,
  klass,
  date = NULL,
  variant = NULL,
  correspond = NULL,
  language = "nb",
  output_level = NULL,
  output = "name",
  format = TRUE
)

Arguments

Value

A vector or data frame is returned with names and/or code of the desired output level.

Examples

data(klassdata)
kommune_names <- apply_klass(
  x = klassdata$kommune,
  classification = 131,
  language = "en",
  format = FALSE
)

Description

Correspondence list Print a list of correspondence tables for a given classification with source and target IDs

Usage

correspond_list(classification, date = NULL)

CorrespondList(klass, date = NULL)

Arguments

Value

Data frame with list of corrsepondence tables, source ID and target ID.

Examples

correspond_list("7")

Description

Find which codes should be combined to reconstruct older or newer codes across multiple dates. Creates groups dynamically to fit the provided date range, with flexible labelling functionality.

Usage

find_equivalent_codes(
  classification,
  dates,
  labels = TRUE,
  graph = klass_graph(classification),
  date_format = "%Y"
)

find_equivalents(
  classification,
  dates,
  labels = TRUE,
  graph = klass_graph(classification),
  date_format = "%Y"
)

Arguments

"1508 Ålesund, 1580 Haram"

list(
  label1 = function(code, name, date, ...) {
    label_codes <- date == max(date)
    paste(code[label_codes], name[label_codes], collapse = ", ")
  },
  label2 = function(code, date, ...) {
    label_codes <- date == max(date)
    paste(code[label_codes], collapse = ", ")
  }
)

Details

This function provides a solution to the problem of split or combined codes in Klass classifications. When using update_klass to ask "what is this code in this version of the classification in this other version of the classification?", the answer is sometimes that the code has been split into two or more codes (or combined from two or more codes, if trying to back-date a code), and therefore that the code cannot be updated.

The solution provided by find_equivalent_codes is answering the question: "in these versions of the classification, which codes were equivalent to this code in this other version of the classification?".

Consider the following example of two codes combining into one. Here, "a" and "b" are valid at t1, and are combined into "c" at t2.

t1     t2
a ──┰─> c
    ┃
b ──┚

update_klass would inform us that "a" can be updated to "c" at t2, unless we specified combine = FALSE, in which case the result would be NA. find_equivalent_codes() would inform us that the equivalent of the codes "a" and "b" in t1 at t2 is "c".

We can also consider a code splitting into two. In this example, "a" is valid at t1, and splits into "b" and "c" at t2.

t1     t2
a
├─────> b
└─────> c

update_klass is unable to provide an updated code due to the split, and would return NA. find_equivalent_codes would inform us that the equivalent codes of "a" at t1 is "b" and "c" at t2.

find_equivalent_codes can handle more than two dates. In the following example, "a" splits into "b" and "c" at t2, and "b" and "c" combine into "d" at t3. find_equivalent_codes can inform us that "a" is equivalent to "b" and "c" at t2, and "d" at t3.

t1     t2     t3
a
├─────> b ┐
└─────> c ┴─> d

find_equivalent_codes will only search in the time range we specify. As a consequence, generating sets of equivalent codes over longer time spans will generally create larger sets than using shorter time spans.

To illustrate this behavior, we can add a new code "e" to the previous example, and have "d" and "e" combine into "f" at t4.

t1     t2     t3     t4
a
├─────> b ┐
└─────> c ┴─> d ┐
e ──────────────┴──> f

Finding the equivalents of "a" in t1 at t2 and t3 returns the same sets as before:

• t1: "a"

• t2: "b" and "c"

• t3: "d"

However, if we also wanted to know the equivalent set for t4, the result would be:

• t1: "a" and "e"

• t2: "b", "c" and "e"

• t3: "d" and "e"

• t4: "f"

find_equivalents is a legacy alias for find_equivalent_codes.

Value

A data.frame with columns:

• date containing the input dates

• code containing the set of equivalent codes in each date

• name containing the names of each code

• validFrom and validTo values for each code returned

• By default, labels, giving a unique group label for each group of equivalent sets.

Description

Find the equivalent sets of a node at various dates

Usage

find_equivalent_nodes(node, dates, graph)

Arguments

Value

A named list of length(dates) containing the equivalent nodes for each date. The names of the returned list are the dates provided in date, coerced to character.

Description

Identify corresponding family from a classification number

Usage

get_family(classification)

GetFamily(klass)

Arguments

Value

Family number

Examples

get_family(classification = 7)

Description

Fetch Statistics Norway classification data using API

Usage

get_klass(
  classification,
  date = NULL,
  correspond = NULL,
  correspondID = NULL,
  variant = NULL,
  output_level = NULL,
  language = "nb",
  output_style = "normal",
  notes = FALSE,
  quiet = TRUE
)

GetKlass(
  klass,
  date = NULL,
  correspond = NULL,
  correspondID = NULL,
  variant = NULL,
  output_level = NULL,
  language = "nb",
  output_style = "normal",
  notes = FALSE,
  quiet = TRUE
)

Arguments

Value

The function returns a data frame of the specified classification/correspondence table. Output variables include: code, parentCode, level, and name for standard lists. For correspondence tables variables include: sourceCode, sourceName, targetCode and targetName. For date correspondence tables variables include: oldCode, oldName, newCode and newName. For "wide" output, code and name with level suffixes is specified. For date ranges, validFromInRequestedRange and validToInRequestedRange give the dates for the classification. Variable ChangeOccured gives the effective date for classification change in classification change tables.

Examples

# Get classification for occupation classifications
head(get_klass(classification = "7"))
# Get classification for occupation classifications in English
head(get_klass(classification = "7", language = "en"))

Description

Get the name of a classification version

Usage

get_name(version)

GetName(version)

Arguments

Value

string or vector of strings with name of version

Examples

get_name("33")

Description

Get version number of a class given a date

Usage

get_version(classification = NULL, date = NULL, family = NULL, klassNr = FALSE)

GetVersion(klass = NULL, date = NULL, family = NULL, klassNr = FALSE)

Arguments

Value

Number, vector or data frame with version numbers and calssification numbers if specified.

Examples

get_version(7)

Description

A nested list of graph data for using in testing

Usage

klass_131_1964_graph

Format

An object of class igraph of length 2000.

Description

A nested list of graph data for using in testing

Usage

klass_131_2020_graph

Format

An object of class igraph of length 2000.

Description

A nested list of graph data for using in testing

Usage

klass_131_graph

Format

An object of class igraph of length 2000.

Description

Build a directed graph of code changes based on a Klass classification

Usage

klass_graph(classification, date = NULL)

Arguments

Value

An igraph object with the vertexes representing codes, and edges representing changes between codes. The direction of the edges represent changes towards the date specified in date.

Examples

library(klassR)

# Build a graph directed towards the most recent codes
## Not run: 
klass_131 <- klass_graph(131)

## End(Not run)

# Build a graph directed towards valid codes in 2020.
## Not run: 
klass_131_2020 <- klass_graph(131, "2020-01-01")

## End(Not run)

Description

Given a Klass graph, find the node corresponding to a code and (optionally) a date.

Usage

klass_node(graph, x, date = NA)

Arguments

Value

The node in the graph corresponding to the supplied code. If date is not provided, the node with the most recent code is returned. If date is provided, the code with date between validFrom and validTo is returned.

Examples

## Not run: 
# Build a graph directed towards the most recent codes.
library(klassR)
klass_131 <- klass_graph(131)

# Find the most recent node in the graph representing the code "0101" (Halden,
# valid to 2020.)
halden_node <- klass_node(klass_131, "0101")

## End(Not run)

Description

A dataset containing variables for testing of Statistics Norways classification API with the klassR package. Some observations are missing or incorrect for testing and demonstrations.

Usage

klassdata

Format

A data frame containing 100 rows and 7 variables:

ID

Identification number

sex

1/2 variable for sex

education

4-digit number for education standard ISCED97 (level and subject area) NUS (klass = 66) 2015.01.01

kommune

4-digit code for Norwegian municipality (klass = 131). Based on 2015.01.01

kommune2

Numeric variable for Norwegian municipality with dropped leading zero's for testing (klass = 131). Based on 2015.01.01

nace5

5-digit code for industry (NACE). Based on 01.01.2015 standard industry codes (klass = 7)

occupation

4-digit occupation codes using standard for STYRK-08 (klass = 7) 2015.01.01

Description

Classification family list Print a list of all families and the number of classifications in each

Usage

list_family(family = NULL, codelists = FALSE, language = "nb")

ListFamily(family = NULL, codelists = FALSE, language = "nb")

Arguments

Value

dataset containing a list of families

Examples

list_family(family = 1)

Description

Classification list Get a full list of all classifications and codelists

Usage

list_klass(codelists = FALSE, language = "nb")

ListKlass(codelists = FALSE, language = "nb")

Arguments

Value

A data frame containing a full list of classifications. The data frame includes the classification name, number, family and type.

Examples

head(list_klass(codelists = TRUE))

Description

Search Klass

Usage

search_klass(query, codelists = FALSE, size = 20)

SearchKlass(query, codelists = FALSE, size = 20)

Arguments

Value

Data frame of possible classifications that match the query

Examples

search_klass("occupation")

Description

Update multiple Klass codes to a desired date.

Usage

update_klass(
  codes,
  dates = NA,
  classification = NULL,
  date = NULL,
  graph = klass_graph(classification, date),
  output = "code",
  report = FALSE,
  combine = TRUE
)

Arguments

Value

If output = "code", a vector of length length(codes) containing either a code if the update is successful or NA if the code has been split. If combine = FALSE, a code being combined with another code will also return NA.

If output == TRUE, a list of length length(codes) containing data.frames detailing the codes visited through the node search. The tables have the following columns.

If report == TRUE and length(output) > 1 | TRUE, the result will be a list of data.frames with number of rows equal to the number of codes in the sequence of changes between the input codes and output codes. The columns in the data.frames are specified with output.

If report == TRUE and length(output) == 1, the result will be a list of character vectors with length equal to the number of codes in the sequence of changes between the input code and output code. The contents of the character vectors is specified with output.

If report == FALSE and length(output) > 1 | TRUE the result will be a list of data.frames with one row representing the last code in the change sequence and columns specified by output. If a code has been split, the result will be NA. If combine == FALSE and a code is the result of a combination of codes, the result will be NA.

If report == FALSE and length(output) == 1, the result will be a character vector containing information about the updated codes specified by output. If a code has been split, the result will be NA. If combine == FALSE and a code is the result of a combination of codes, the result will be NA.

Examples

library(klassR)
codes <- get_klass(131, date = "2020-01-01")[["code"]]

## Not run: 
updated_codes <- update_klass(codes,
  dates = "2020-01-01",
  classification = 131
)

## End(Not run)

Description

Given a node and a graph, find the node at the end of a sequence of changes.

Usage

update_klass_node(graph, node)

Arguments

Value

A sequence of vertices, starting with node and ending with the last visited node.

Examples

## Not run: 
# Build a graph directed towards the most recent codes.
library(klassR)
klass_131 <- klass_graph(131)

# Find the most recent node in the graph representing the code "0101" (Halden,
# valid to 2020.)
halden_node <- klass_node(klass_131, "0101")

# Find the most recent code corresponding to 0101 Halden
halden_node_updated <- update_klass_node(klass_131, halden_node)

## End(Not run)