Nothing
R/r6.R
In fio: Friendly Input-Output Analysis
#' @title
#' R6 class for input-output matrix
#'
#' @description
#' R6 class for input-output matrix.
#'
#' @param id (`character`)\cr
#' Identifier for the input-output matrix.
#' @param intermediate_transactions (`matrix`)\cr
#' Intermediate transactions matrix.
#' @param total_production (`matrix`)\cr
#' Total production vector.
#' @param household_consumption (`matrix`)\cr
#' Household consumption vector.
#' @param government_consumption (`matrix`)\cr
#' Government consumption vector.
#' @param exports (`matrix`)\cr
#' Exports vector.
#' @param final_demand_others (`matrix`)\cr
#' Other vectors of final demand that doesn't have dedicated slots.
#' Setting column names is advised for better readability.
#' @param imports (`matrix`)\cr
#' Imports vector.
#' @param taxes (`matrix`)\cr
#' Taxes vector.
#' @param wages (`matrix`)\cr
#' Wages vector.
#' @param operating_income (`matrix`)\cr
#' Operating income vector.
#' @param value_added_others (`matrix`)\cr
#' Other vectors of value-added that doesn't have dedicated slots.
#' Setting row names is advised for better readability.
#' @param occupation (`matrix`)\cr
#' Occupation matrix.
#'
#' @return A new instance of the `iom` class.
#'
#' @examples
#' # data
#' intermediate_transactions <- matrix(c(1, 2, 3, 4, 5, 6, 7, 8, 9), 3, 3)
#' total_production <- matrix(c(100, 200, 300), 1, 3)
#' exports <- matrix(c(10, 20, 30), 3, 1)
#' households <- matrix(as.numeric(4:6), 3, 1)
#' imports <- matrix(c(5, 10, 15), 1, 3)
#' jobs <- matrix(c(10, 12, 15), 1, 3)
#' taxes <- matrix(c(2, 5, 10), 1, 3)
#' wages <- matrix(c(11, 12, 13), 1, 3)
#'
#' # a new iom instance can be created by passing just intermediate transactions and total production
#' my_iom <- iom$new(
#' "example_1",
#' intermediate_transactions,
#' total_production
#' )
#'
#' # or by passing optional arguments
#' my_iom <- iom$new(
#' "example_2",
#' intermediate_transactions,
#' total_production,
#' household_consumption = households,
#' exports = exports,
#' imports = imports,
#' taxes = taxes,
#' wages = wages,
#' occupation = jobs
#' )
#'
#' @importFrom Rdpack reprompt
#' @import R6
#' @export
# input-output matrix class
iom <- R6Class(
classname = "iom",
public = list(
#' @field id (`character`)\cr
#' Identifier of the new instance.
id = NULL,
#' @field intermediate_transactions (`matrix`)\cr
#' Intermediate transactions matrix.
intermediate_transactions = NULL,
#' @field total_production (`matrix`)\cr
#' Total production vector.
total_production = NULL,
#' @field household_consumption (`matrix`)\cr
#' Household consumption vector.
household_consumption = NULL,
#' @field government_consumption (`matrix`)\cr
#' Government consumption vector.
government_consumption = NULL,
#' @field exports (`matrix`)\cr
#' Exports vector.
exports = NULL,
#' @field final_demand_others (`matrix`)\cr
#' Other vectors of final demand that doesn't have dedicated slots.
final_demand_others = NULL,
#' @field final_demand_matrix (`matrix`)\cr
#' Aggregates final demand vectors into a matrix.
final_demand_matrix = NULL,
#' @field imports (`matrix`)\cr
#' Imports vector.
imports = NULL,
#' @field taxes (`matrix`)\cr
#' Taxes vector.
taxes = NULL,
#' @field wages (`matrix`)\cr
#' Wages vector.
wages = NULL,
#' @field operating_income (`matrix`)\cr
#' Operating income vector.
operating_income = NULL,
#' @field value_added_others (`matrix`)\cr
#' Other vectors of value-added that doesn't have dedicated slots.
value_added_others = NULL,
#' @field value_added_matrix (`matrix`)\cr
#' Aggregates value-added vectors into a matrix.
value_added_matrix = NULL,
#' @field occupation (`matrix`)\cr
#' Occupation vector.
occupation = NULL,
#' @field technical_coefficients_matrix (`matrix`)\cr
#' Technical coefficients matrix.
technical_coefficients_matrix = NULL,
#' @field leontief_inverse_matrix (`matrix`)\cr
#' Leontief inverse matrix.
leontief_inverse_matrix = NULL,
#' @field multiplier_output (`data.frame`)\cr
#' Output multiplier dataframe.
multiplier_output = NULL,
#' @field multiplier_employment (`data.frame`)\cr
#' Employment multiplier dataframe.
multiplier_employment = NULL,
#' @field multiplier_taxes (`data.frame`)\cr
#' Taxes multiplier dataframe.
multiplier_taxes = NULL,
#' @field multiplier_wages (`data.frame`)\cr
#' Wages multiplier dataframe.
multiplier_wages = NULL,
#' @field field_influence (`matrix`)\cr
#' Influence field matrix.
field_influence = NULL,
#' @field key_sectors (`data.frame`)\cr
#' Key sectors dataframe.
key_sectors = NULL,
#' @field allocation_coefficients_matrix (`matrix`)\cr
#' Allocation coefficients matrix.
allocation_coefficients_matrix = NULL,
#' @field ghosh_inverse_matrix (`matrix`)\cr
#' Ghosh inverse matrix.
ghosh_inverse_matrix = NULL,
#' @field hypothetical_extraction (`matrix`)\cr
#' Absolute and relative backward and forward differences in total output after a hypothetical extraction
hypothetical_extraction = NULL,
#' @description
#' Creates a new instance of this [R6][R6::R6Class] class.
initialize = function(id,
intermediate_transactions,
total_production,
household_consumption = NULL,
government_consumption = NULL,
exports = NULL,
final_demand_others = NULL,
imports = NULL,
taxes = NULL,
wages = NULL,
operating_income = NULL,
value_added_others = NULL,
occupation = NULL) {
### assertions ###
# check class
for (matrix in private$iom_elements()) {
if (!is.null(get_var(matrix)) && !is.matrix(get_var(matrix))) {
cli::cli_h1("Error in matrix class")
alert(paste("Try coerce", matrix, "to a matrix using as.matrix() function."))
error(paste(matrix, "must be a matrix."))
}
}
# check dimensions
if (nrow(intermediate_transactions) != ncol(intermediate_transactions)) {
cli::cli_h1("Error in matrix dimensions")
error("intermediate_transactions must be a square matrix")
}
for (matrix in c(
"household_consumption",
"government_consumption",
"exports",
"final_demand_others"
)) {
if (
!is.null(get_var(matrix))
&& nrow(get_var(matrix)) != nrow(intermediate_transactions)
) {
cli::cli_h1("Error in matrix dimensions")
error(
paste(
matrix,
"must have the same number of rows than `intermediate_transactions`,
which is",
nrow(intermediate_transactions),
"rows. But has",
nrow(get_var(matrix)),
"rows."
)
)
}
}
for (matrix in c(
"imports",
"taxes",
"wages",
"operating_income",
"value_added_others",
"occupation",
"total_production"
)) {
if (!is.null(get_var(matrix)) && ncol(get_var(matrix)) != ncol(intermediate_transactions)) {
cli::cli_h1("Error in matrix dimensions")
error(
paste(
matrix,
"must have the same number of columns than `intermediate_transactions`, which is",
ncol(intermediate_transactions),
"columns. But",
matrix,
"has",
ncol(get_var(matrix)),
"columns."
)
)
}
}
# check number format
for (matrix in private$iom_elements()) {
if (!is.null(get_var(matrix))) {
# Check if the matrix storage mode is not double
if (storage.mode(get_var(matrix)) != "double") {
cli::cli_h1("Error in matrix number format")
alert(
paste(
"Try coerce",
matrix,
"elements to double using as.numeric()."
)
)
error(paste(matrix, "elements must be of type double."))
}
}
}
# set data members
self$id <- id
self$intermediate_transactions <- intermediate_transactions
self$total_production <- total_production
self$household_consumption <- set_colnames(household_consumption)
self$government_consumption <- set_colnames(government_consumption)
self$exports <- set_colnames(exports)
self$final_demand_others <- final_demand_others
self$imports <- set_rownames(imports)
self$taxes <- set_rownames(taxes)
self$wages <- set_rownames(wages)
self$operating_income <- set_rownames(operating_income)
self$value_added_others <- value_added_others
self$occupation <- set_rownames(occupation)
},
#' @description
#' Adds a `matrix` to the `iom` object.
#' @param matrix_name (`character`)\cr
#' One of household_consumption, government_consumption, exports, final_demand_others,
#' imports, taxes, wages, operating income, value_added_others or occupation matrix to be added.
#' @param matrix (`matrix`)\cr
#' Matrix object to be added.
#' @return
#' Self (invisibly).
#' @examples
#' # data
#' intermediate_transactions <- matrix(c(1, 2, 3, 4, 5, 6, 7, 8, 9), 3, 3)
#' total_production <- matrix(c(100, 200, 300), 1, 3)
#' # instantiate iom object
#' my_iom <- iom$new("mock", intermediate_transactions, total_production)
#' # Create a dummy matrix
#' exports_data <- matrix(as.numeric(1:3), 3, 1)
#' # Add the matrix
#' my_iom$add("exports", exports_data)
add = function(matrix_name, matrix) {
# check arg
choices <- private$iom_elements()
if (!matrix_name %in% choices) {
cli::cli_h1("Error in matrix_name")
error(paste("matrix_name must be one of", paste(choices, collapse = ", ")))
}
# check class
if (!is.matrix(matrix)) {
cli::cli_h1("Error in matrix class")
alert(paste("Try coerce", matrix_name, "to a matrix using as.matrix() function."))
error(paste(matrix_name, "must be a matrix."))
}
# check dimensions
if (matrix_name %in% c("household_consumption", "government_consumption", "exports", "final_demand_others")) {
if (nrow(matrix) != nrow(self$intermediate_transactions)) {
cli::cli_h1("Error in matrix dimensions")
error(
paste(
matrix_name,
"must have the same number of rows than intermediate_transactions, which is",
nrow(self$intermediate_transactions),
"rows. But",
matrix_name,
"has",
nrow(matrix),
"rows."
)
)
}
} else {
if (ncol(matrix) != ncol(self$intermediate_transactions)) {
cli::cli_h1("Error in matrix dimensions")
error(
paste(
matrix_name,
"must have the same number of columns than intermediate_transactions, which is",
ncol(self$intermediate_transactions),
"columns. But", matrix_name, "has", ncol(matrix), "columns."
)
)
}
}
# import matrix
self[[matrix_name]] <- matrix
invisible(self)
},
#' @description
#' Removes a `matrix` from the `iom` object.
#' @param matrix_name (`character`)\cr
#' One of household_consumption, government_consumption, exports, final_demand_others,
#' imports, taxes, wages, operating_income, value_added_others or occupation matrix to be removed.
#' @return Self (invisibly).
#' @examples
#' # data
#' intermediate_transactions <- matrix(c(1, 2, 3, 4, 5, 6, 7, 8, 9), 3, 3)
#' total_production <- matrix(c(100, 200, 300), 1, 3)
#' exports_data <- matrix(as.numeric(1:3), 3, 1)
#' # instantiate iom object
#' my_iom <- iom$new("mock", intermediate_transactions, total_production, exports = exports_data)
#' # Remove the matrix
#' my_iom$remove("exports")
remove = function(matrix_name) {
# check arg
choices <- private$iom_elements()
if (!matrix_name %in% choices) {
cli::cli_h1("Error in matrix_name")
error(paste("matrix_name must be one of", paste(choices, collapse = ", ")))
}
# remove matrix
self[[matrix_name]] <- NULL
invisible(self)
},
#' @description
#' Aggregates final demand vectors into the `final_demand_matrix` field.
#' @details
#' Some methods, as `$compute_hypothetical_extraction()`, require the final demand and value-added vectors
#' to be aggregated into a matrix. This method does this aggregation, binding the vectors into
#' `$final_demand_matrix`.
#' @return
#' This functions doesn't returns a value.
#' @examples
#' # data
#' intermediate_transactions <- matrix(c(1, 2, 3, 4, 5, 6, 7, 8, 9), 3, 3)
#' total_production <- matrix(c(100, 200, 300), 1, 3)
#' exports_data <- matrix(c(10, 20, 30), 3, 1)
#' households <- matrix(as.numeric(4:6), 3, 1)
#' # instantiate iom object
#' my_iom <- iom$new(
#' "mock",
#' intermediate_transactions,
#' total_production,
#' exports = exports_data,
#' household_consumption = households
#' )
#' # aggregate all final demand vectors
#' my_iom$update_final_demand_matrix()
#' # check final demand matrix
#' my_iom$final_demand_matrix
update_final_demand_matrix = function() {
# bind final demand vectors
self$final_demand_matrix <- as.matrix(cbind(
self$household_consumption,
self$government_consumption,
self$exports,
self$final_demand_others
))
},
#' @description
#' Aggregates value-added vectors into the `value_added_matrix` field.
#' @details
#' Some methods, as `$compute_hypothetical_extraction()`, require the final demand and value-added vectors
#' to be aggregated into a matrix. This method does this aggregation, binding the vectors into
#' `$value_added_matrix`.
#' @return
#' This functions doesn't returns a value.
#' @examples
#' # data
#' intermediate_transactions <- matrix(c(1, 2, 3, 4, 5, 6, 7, 8, 9), 3, 3)
#' total_production <- matrix(c(100, 200, 300), 1, 3)
#' imports_data <- matrix(c(5, 10, 15), 1, 3)
#' taxes_data <- matrix(c(2, 5, 10), 1, 3)
#' # instantiate iom object
#' my_iom <- iom$new(
#' "mock",
#' intermediate_transactions,
#' total_production,
#' imports = imports_data,
#' taxes = taxes_data
#' )
#' # aggregate all value-added vectors
#' my_iom$update_value_added_matrix()
#' # check value-added matrix
#' my_iom$value_added_matrix
update_value_added_matrix = function() {
# bind value-added vectors
self$value_added_matrix <- as.matrix(rbind(
self$imports,
self$taxes,
self$wages,
self$operating_income,
self$value_added_others
))
},
#' @description
#' Computes the technical coefficients matrix and populate the `technical_coefficients_matrix` field with the
#' resulting `(matrix)`.
#' @details
#' It computes the technical coefficients matrix, a \eqn{n x n} matrix known as `A` matrix which is the column-wise
#' ratio of intermediate transactions to total production \insertCite{leontief_economia_1983}{fio}.
#'
#' ## References
#' \insertCited{}
#' @return
#' Self (invisibly).
#' @examples
#' intermediate_transactions <- matrix(c(1, 2, 3, 4, 5, 6, 7, 8, 9), 3, 3)
#' total_production <- matrix(c(100, 200, 300), 1, 3)
#' # instantiate iom object
#' my_iom <- iom$new("test", intermediate_transactions, total_production)
#' # Calculate the technical coefficients
#' my_iom$compute_tech_coeff()
#' # show the technical coefficients
#' my_iom$technical_coefficients_matrix
compute_tech_coeff = function() {
# save row and column names
row_names <- if (is.null(rownames(self$intermediate_transactions))) {
seq_len(nrow(self$intermediate_transactions))
} else {
rownames(self$intermediate_transactions)
}
col_names <- if (is.null(colnames(self$intermediate_transactions))) {
seq_len(ncol(self$intermediate_transactions))
} else {
colnames(self$intermediate_transactions)
}
# calculate technical coefficients matrix
technical_coefficients_matrix <- compute_tech_coeff(
intermediate_transactions = self$intermediate_transactions,
total_production = self$total_production
)
# set row and column names
rownames(technical_coefficients_matrix) <- row_names
colnames(technical_coefficients_matrix) <- col_names
# store matrix
self$technical_coefficients_matrix <- technical_coefficients_matrix
invisible(self)
},
#' @description
#' Computes the Leontief inverse matrix and populate the `leontief_inverse_matrix` field with the resulting
#' `(matrix)`.
#' @details
#' It computes the Leontief inverse matrix \insertCite{leontief_economia_1983}{fio}, which is the inverse of the
#' Leontief matrix, defined as:
#'
#' \deqn{L = I - A}
#'
#' where I is the identity matrix and A is the technical coefficients matrix.
#' The Leontief inverse matrix is calculated by solving the following equation:
#'
#' \deqn{L^{-1} = (I - A)^{-1}}
#'
#' Since the Leontief matrix is a square matrix and the subtraction of the technical coefficients matrix from the
#' identity matrix guarantees that the Leontief matrix is invertible, underlined Rust function uses LU decomposition
#' to solve the equation.
#'
#' ## References:
#' \insertCited{}
#' @return
#' Self (invisibly).
#' @examples
#' intermediate_transactions <- matrix(c(1, 2, 3, 4, 5, 6, 7, 8, 9), 3, 3)
#' total_production <- matrix(c(100, 200, 300), 1, 3)
#' # instantiate iom object
#' my_iom <- fio::iom$new("test", intermediate_transactions, total_production)
#' # calculate the technical coefficients
#' my_iom$compute_tech_coeff()
#' # calculate the Leontief inverse
#' my_iom$compute_leontief_inverse()
#' # show the Leontief inverse
#' my_iom$leontief_inverse_matrix
compute_leontief_inverse = function() {
# check if technical coefficients matrix is available
if (is.null(self$technical_coefficients_matrix)) {
cli::cli_h1("Error in technical_coefficients_matrix")
error("You must compute the technical coefficients matrix first. Run compute_tech_coeff() method.")
}
# save row and column names
row_names <- rownames(self$technical_coefficients_matrix)
col_names <- colnames(self$technical_coefficients_matrix)
# computes leontief inverse matrix
leontief_inverse_matrix <- compute_leontief_inverse(
tech_coeff = self$technical_coefficients_matrix
)
# set row and column names
rownames(leontief_inverse_matrix) <- row_names
colnames(leontief_inverse_matrix) <- col_names
# store matrix
self$leontief_inverse_matrix <- leontief_inverse_matrix
invisible(self)
},
#' @description
#' Computes the output multiplier and populate the `multiplier_output` field with the resulting `(data.frame)`.
#' @details
#' An output multiplier for sector *j* is defined as the total value of production in all sectors of the economy
#' that is necessary in order to satisfy a monetary unit (e.g., a dollar) worth of final demand for sector *j*'s
#' output \insertCite{miller_input-output_2009}{fio}.
#'
#' This method computes the simple output multiplier, defined as the column sums of the Leontief inverse matrix,
#' the direct and indirect output multipliers, which are the column sums of the technical
#' coefficients matrix and the difference between total and direct output multipliers, respectively
#' \insertCite{vale_alise_2020}{fio}.
#'
#' ## References
#' \insertCited{}
#' @return
#' Self (invisibly).
#' @examples
#' # data
#' intermediate_transactions <- matrix(c(1, 2, 3, 4, 5, 6, 7, 8, 9), 3, 3)
#' total_production <- matrix(c(100, 200, 300), 1, 3)
#' # instantiate iom object
#' my_iom <- fio::iom$new("test", intermediate_transactions, total_production)
#' # calculate the technical coefficients
#' my_iom$compute_tech_coeff()
#' # calculate the Leontief inverse
#' my_iom$compute_leontief_inverse()
#' # calculate the output multiplier
#' my_iom$compute_multiplier_output()
#' # show the output multiplier
#' my_iom$multiplier_output
compute_multiplier_output = function() {
# check if leontief inverse matrix is available
if (is.null(self$leontief_inverse_matrix)) {
cli::cli_h1("Error in leontief_inverse_matrix")
error("You must compute the leontief inverse matrix first. Run compute_leontief_inverse() method.")
}
# save column names
col_names <- colnames(self$leontief_inverse_matrix)
# compute output multiplier vector
multiplier_output_simple <- compute_multiplier_output(
leontief_inverse_matrix = self$leontief_inverse_matrix
)
# compute direct output multiplier vector
multiplier_output_direct <- compute_multiplier_output_direct(
technical_coefficients_matrix = self$technical_coefficients_matrix
)
# compute indirect output multiplier vector
multiplier_output_indirect <- compute_multiplier_output_indirect(
technical_coefficients_matrix = self$technical_coefficients_matrix,
leontief_inverse_matrix = self$leontief_inverse_matrix
)
multiplier_output <- data.frame(
sector = col_names,
multiplier_simple = multiplier_output_simple,
multiplier_direct = multiplier_output_direct,
multiplier_indirect = multiplier_output_indirect
)
# store vector
self$multiplier_output <- multiplier_output
invisible(self)
},
#' @description
#' Computes the employment multiplier and populate the `multiplier_employment` field with the resulting
#' `(data.frame)`.
#' @details
#' The employment multiplier for sector *j* relates the jobs created in each sector in response to a
#' initial exogenous shock \insertCite{miller_input-output_2009}{fio}.
#'
#' Current implementation follows \insertCite{vale_alise_2020}{fio}.
#'
#' ## References
#' \insertCited{}
#' @return
#' Self (invisibly).
#' @examples
#' # data
#' intermediate_transactions <- matrix(c(1, 2, 3, 4, 5, 6, 7, 8, 9), 3, 3)
#' total_production <- matrix(c(100, 200, 300), 1, 3)
#' jobs_data <- matrix(c(10, 12, 15), 1, 3)
#' # instantiate iom object
#' my_iom <- fio::iom$new("test", intermediate_transactions, total_production, occupation = jobs_data)
#' # calculate the technical coefficients
#' my_iom$compute_tech_coeff()
#' # calculate the Leontief inverse
#' my_iom$compute_leontief_inverse()
#' # calculate the employment multiplier
#' my_iom$compute_multiplier_employment()
#' # show the employment multiplier
#' my_iom$multiplier_employment
compute_multiplier_employment = function() {
# check if leontief inverse matrix is available
if (is.null(self$leontief_inverse_matrix)) {
cli::cli_h1("Error in leontief_inverse_matrix")
error("You must compute the leontief inverse matrix first. Run compute_leontief_inverse() method.")
}
# save column names
col_names <- colnames(self$leontief_inverse_matrix)
# compute employment requirements
employment_requirements <- compute_requirements_value_added(
value_added_element = self$occupation,
total_production = self$total_production
)
# compute employment multiplier vector
multiplier_employment_simple <- compute_multiplier_value_added(
value_added_requirements = employment_requirements,
leontief_inverse_matrix = self$leontief_inverse_matrix
)
# compute indirect employment multiplier
multiplier_employment_indirect <- compute_multiplier_value_added_indirect(
value_added_element = self$occupation,
total_production = self$total_production,
leontief_inverse_matrix = self$leontief_inverse_matrix
)
multiplier_employment <- data.frame(
sector = col_names,
multiplier_simple = multiplier_employment_simple,
multiplier_direct = employment_requirements,
multiplier_indirect = multiplier_employment_indirect
)
# store vector
self$multiplier_employment <- multiplier_employment
invisible(self)
},
#' @description
#' Computes the wages multiplier dataframe and populate the `multiplier_wages` field with the resulting
#' `(data.frame)`.
#' @details
#' The wages multiplier for sector *j* relates increases in wages for each
#' sector in response to a initial exogenous shock
#' \insertCite{miller_input-output_2009}{fio}.
#'
#' Current implementation follows \insertCite{vale_alise_2020}{fio}.
#'
#' ## References
#' \insertCited{}
#' @return
#' Self (invisibly).
#' @examples
#' # data
#' intermediate_transactions <- matrix(c(1, 2, 3, 4, 5, 6, 7, 8, 9), 3, 3)
#' total_production <- matrix(c(100, 200, 300), 1, 3)
#' wages_data <- matrix(c(10, 12, 15), 1, 3)
#' # instantiate iom object
#' my_iom <- fio::iom$new("test", intermediate_transactions, total_production, wages = wages_data)
#' # calculate the technical coefficients
#' my_iom$compute_tech_coeff()
#' # calculate the Leontief inverse
#' my_iom$compute_leontief_inverse()
#' # calculate the wages multiplier
#' my_iom$compute_multiplier_wages()
#' # show the wages multiplier
#' my_iom$multiplier_wages
compute_multiplier_wages = function() {
# check if leontief inverse matrix is available
if (is.null(self$leontief_inverse_matrix)) {
cli::cli_h1("Error in leontief_inverse_matrix")
error("You must compute the leontief inverse matrix first. Run compute_leontief_inverse() method.")
}
# save column names
col_names <- colnames(self$leontief_inverse_matrix)
# compute wages requirements
wages_requirements <- compute_requirements_value_added(
value_added_element = self$wages,
total_production = self$total_production
)
# compute wages multiplier vector
multiplier_wages_simple <- compute_multiplier_value_added(
value_added_requirements = wages_requirements,
leontief_inverse_matrix = self$leontief_inverse_matrix
)
# compute indirect wages multiplier
multiplier_wages_indirect <- compute_multiplier_value_added_indirect(
value_added_element = self$wages,
total_production = self$total_production,
leontief_inverse_matrix = self$leontief_inverse_matrix
)
multiplier_wages <- data.frame(
sector = col_names,
multiplier_simple = multiplier_wages_simple,
multiplier_direct = wages_requirements,
multiplier_indirect = multiplier_wages_indirect
)
# store vector
self$multiplier_wages <- multiplier_wages
invisible(self)
},
#' @description
#' Computes the taxes multiplier and populate the `multiplier_taxes` field with
#' the resulting `(data.frame)`.
#' @details
#' The taxes multiplier for sector *j* relates the increases on tax revenue from
#' each sector in response to a initial exogenous shock
#' \insertCite{miller_input-output_2009}{fio}.
#'
#' Current implementation follows \insertCite{vale_alise_2020}{fio}.
#'
#' ## References
#' \insertCited{}
#' @return
#' Self (invisibly).
#' @examples
#' # data
#' intermediate_transactions <- matrix(c(1, 2, 3, 4, 5, 6, 7, 8, 9), 3, 3)
#' total_production <- matrix(c(100, 200, 300), 1, 3)
#' tax_data <- matrix(c(10, 12, 15), 1, 3)
#' # instantiate iom object
#' my_iom <- fio::iom$new("test", intermediate_transactions, total_production, taxes = tax_data)
#' # calculate the technical coefficients
#' my_iom$compute_tech_coeff()
#' # calculate the Leontief inverse
#' my_iom$compute_leontief_inverse()
#' # calculate the tax multiplier
#' my_iom$compute_multiplier_taxes()
#' # show the taxes multiplier
#' my_iom$multiplier_taxes
compute_multiplier_taxes = function() {
# check if leontief inverse matrix is available
if (is.null(self$leontief_inverse_matrix)) {
cli::cli_h1("Error in leontief_inverse_matrix")
error("You must compute the leontief inverse matrix first. Run compute_leontief_inverse() method.")
}
# save column names
col_names <- colnames(self$leontief_inverse_matrix)
# compute taxes requirements
taxes_requirements <- compute_requirements_value_added(
value_added_element = self$taxes,
total_production = self$total_production
)
# compute taxes multiplier vector
multiplier_taxes_simple <- compute_multiplier_value_added(
value_added_requirements = taxes_requirements,
leontief_inverse_matrix = self$leontief_inverse_matrix
)
# compute indirect taxes multiplier
multiplier_taxes_indirect <- compute_multiplier_value_added_indirect(
value_added_element = self$taxes,
total_production = self$total_production,
leontief_inverse_matrix = self$leontief_inverse_matrix
)
multiplier_taxes <- data.frame(
sector = col_names,
multiplier_simple = multiplier_taxes_simple,
multiplier_direct = taxes_requirements,
multiplier_indirect = multiplier_taxes_indirect
)
# store vector
self$multiplier_taxes <- multiplier_taxes
invisible(self)
},
#' @description
#' Computes the field of influence for all sectors and populate the
#' `field_influence` field with the resulting `(matrix)`.
#' @details
#' The field of influence shows how changes in direct coefficients are
#' distributed throughout the entire economic system, allowing for the
#' determination of which relationships between sectors are most important
#' within the production process.
#'
#' It determines which sectors have the greatest influence over others,
#' specifically, which coefficients, when altered, would have the greatest
#' impact on the system as a whole \insertCite{vale_alise_2020}{fio}.
#'
#' ## References
#' \insertCited{}
#' @param epsilon (`numeric`)\cr
#' Epsilon value. A technical change in the input-output matrix, caused by a variation of size `epsilon` into each
#' element of technical coefficients matrix.
#' @return
#' Self (invisibly).
#' @examples
#' # data
#' intermediate_transactions <- matrix(c(1, 2, 3, 4, 5, 6, 7, 8, 9), 3, 3)
#' total_production <- matrix(c(100, 200, 300), 1, 3)
#' # instantiate iom object
#' my_iom <- fio::iom$new("test", intermediate_transactions, total_production)
#' # calculate the technical coefficients
#' my_iom$compute_tech_coeff()
#' # calculate the Leontief inverse
#' my_iom$compute_leontief_inverse()
#' # calculate field of influence
#' my_iom$compute_field_influence(epsilon = 0.01)
#' # show the field of influence
#' my_iom$field_influence
compute_field_influence = function(epsilon) {
# check if epsilon was set
if (missing(epsilon)) {
cli::cli_h1("Error in epsilon")
error("You must set the epsilon value.")
}
# check if leontief inverse matrix is available
if (is.null(self$leontief_inverse_matrix)) {
cli::cli_h1("Error in leontief_inverse_matrix")
error("You must compute the leontief inverse matrix first. Run compute_leontief_inverse() method.")
}
# save row and column names
row_names <- rownames(self$technical_coefficients_matrix)
col_names <- colnames(self$technical_coefficients_matrix)
# compute influence field matrix
field_influence <- compute_field_influence(
tech_coeff = self$technical_coefficients_matrix,
leontief_inverse_matrix = self$leontief_inverse_matrix,
epsilon = epsilon
)
# set row and column names
rownames(field_influence) <- row_names
colnames(field_influence) <- col_names
# store matrix
self$field_influence <- field_influence
invisible(self)
},
#' @description
#' Computes the key sectors dataframe, based on it's power and sensitivity of dispersion,
#' and populate the `key_sectors` field with the resulting `(data.frame)`.
#' @details
#' Increased production from a sector *j* means that the sector *j* will need to
#' purchase more goods from other sectors. At the same time, it means that more goods from sector *j* will be
#' available for other sectors to purchase. Sectors that are above average in the demand sense (stronger backward
#' linkage) have power of dispersion indices greater than 1. Sectors that are above average in the supply sense
#' (stronger forward linkage) have sensitivity of dispersion indices greater than 1
#' \insertCite{miller_input-output_2009}{fio}.
#'
#' As both power and sensitivity of dispersion are related to average values on the economy, coefficients of
#' variation are also calculated for both indices. The lesser the coefficient of variation, greater the number of
#' sectors on the demand or supply structure of that sector \insertCite{vale_alise_2020}{fio}.
#'
#' ## References
#' \insertCited{}
#' @return
#' Self (invisibly).
#' @examples
#' # data
#' intermediate_transactions <- matrix(c(1, 2, 3, 4, 5, 6, 7, 8, 9), 3, 3)
#' total_production <- matrix(c(100, 200, 300), 1, 3)
#' # instantiate iom object
#' my_iom <- fio::iom$new("test", intermediate_transactions, total_production)
#' # calculate the technical coefficients
#' my_iom$compute_tech_coeff()
#' # calculate the Leontief inverse
#' my_iom$compute_leontief_inverse()
#' # calculate key sectors
#' my_iom$compute_key_sectors()
#' # show the key sectors
#' my_iom$key_sectors
compute_key_sectors = function() {
# check if leontief inverse matrix is available
if (is.null(self$leontief_inverse_matrix)) {
cli::cli_h1("Error in leontief_inverse_matrix")
error("You must compute the leontief inverse matrix first. Run compute_leontief_inverse() method.")
}
# power of dispersion
power_dispersion <- compute_power_dispersion(
leontief_inverse_matrix = self$leontief_inverse_matrix
)
# sensitivity of dispersion
sensitivity_dispersion <- compute_sensitivity_dispersion(
leontief_inverse_matrix = self$leontief_inverse_matrix
)
# power of dispersion coefficients of variation
power_dispersion_cv <- compute_power_dispersion_cv(
leontief_inverse_matrix = self$leontief_inverse_matrix
)
# sensitivity of dispersion coefficients of variation
sensitivity_dispersion_cv <- compute_sensitivity_dispersion_cv(
leontief_inverse_matrix = self$leontief_inverse_matrix
)
# compute key sectors dataframe
key_sectors <- data.frame(
sector = rownames(self$leontief_inverse_matrix),
power_dispersion = power_dispersion,
sensitivity_dispersion = sensitivity_dispersion,
power_dispersion_cv = power_dispersion_cv,
sensitivity_dispersion_cv = sensitivity_dispersion_cv
) |>
within({
key_sectors <- ifelse(sensitivity_dispersion <= 1 & power_dispersion <= 1, "Non-Key Sector", "")
key_sectors <- ifelse(sensitivity_dispersion > 1 & power_dispersion > 1, "Key Sector", key_sectors)
key_sectors <- ifelse(sensitivity_dispersion > 1 & power_dispersion <= 1, "Strong Forward Linkage", key_sectors) # nolint
key_sectors <- ifelse(sensitivity_dispersion <= 1 & power_dispersion > 1, "Strong Backward Linkage", key_sectors) # nolint
})
# store dataframe
self$key_sectors <- key_sectors
invisible(self)
},
#' @description
#' Computes the allocation coefficients matrix and populate the `allocation_coefficients_matrix` field with the
#' resulting `(matrix)`.
#' @details
#' It computes the allocation coefficients matrix, a \eqn{n x n} matrix known as `B` matrix which is the row-wise
#' ratio of intermediate transactions to total production \insertCite{miller_input-output_2009}{fio}.
#'
#' ## References
#' \insertCited{}
#' @return
#' Self (invisibly).
#' @examples
#' intermediate_transactions <- matrix(c(1, 2, 3, 4, 5, 6, 7, 8, 9), 3, 3)
#' total_production <- matrix(c(100, 200, 300), 1, 3)
#' # instantiate iom object
#' my_iom <- fio::iom$new("test", intermediate_transactions, total_production)
#' # Calculate the allocation coefficients
#' my_iom$compute_allocation_coeff()
#' # show the allocation coefficients
#' my_iom$allocation_coefficients_matrix
compute_allocation_coeff = function() {
# save row and column names
row_names <- if (is.null(rownames(self$intermediate_transactions))) {
seq_len(nrow(self$intermediate_transactions))
} else {
rownames(self$intermediate_transactions)
}
col_names <- if (is.null(colnames(self$intermediate_transactions))) {
seq_len(ncol(self$intermediate_transactions))
} else {
colnames(self$intermediate_transactions)
}
# compute allocation coefficients matrix
allocation_coefficients_matrix <- compute_allocation_coeff(
intermediate_transactions = self$intermediate_transactions,
total_production = self$total_production
)
# set row and column names
rownames(allocation_coefficients_matrix) <- row_names
colnames(allocation_coefficients_matrix) <- col_names
# store matrix
self$allocation_coefficients_matrix <- allocation_coefficients_matrix
invisible(self)
},
#' @description
#' Computes the Ghosh inverse matrix and populate the `ghosh_inverse_matrix` field with the resulting `(matrix)`.
#' @details
#' It computes the Ghosh inverse matrix \insertCite{miller_input-output_2009}{fio}, defined as:
#' \deqn{G = (I - B)^{-1}}
#' where I is the identity matrix and B is the allocation coefficients matrix.
#'
#' ## References
#' \insertCited{}
#' @return
#' Self (invisibly).
#' @examples
#' intermediate_transactions <- matrix(c(1, 2, 3, 4, 5, 6, 7, 8, 9), 3, 3)
#' total_production <- matrix(c(100, 200, 300), 1, 3)
#' # instantiate iom object
#' my_iom <- fio::iom$new("test", intermediate_transactions, total_production)
#' # Calculate the allocation coefficients
#' my_iom$compute_allocation_coeff()
#' # Calculate the Ghosh inverse
#' my_iom$compute_ghosh_inverse()
#' # show the Ghosh inverse
#' my_iom$ghosh_inverse_matrix
compute_ghosh_inverse = function() {
# check if allocation coefficients matrix is available
if (is.null(self$allocation_coefficients_matrix)) {
cli::cli_h1("Error in allocation_coefficients_matrix")
error("You must compute the allocation coefficients matrix first. Run compute_allocation_coeff() method.")
}
# save row and column names
row_names <- rownames(self$allocation_coefficients_matrix)
col_names <- colnames(self$allocation_coefficients_matrix)
# compute ghosh inverse matrix
ghosh_inverse_matrix <- compute_ghosh_inverse(
allocation_coeff = self$allocation_coefficients_matrix
)
# set row and column names
rownames(ghosh_inverse_matrix) <- row_names
colnames(ghosh_inverse_matrix) <- col_names
# store matrix
self$ghosh_inverse_matrix <- ghosh_inverse_matrix
invisible(self)
},
#' @description
#' Computes total impact after extracting a each sector and populate the `hypothetical_extraction` field with the
#' resulting `(data.frame)`.
#' @details
#' Computes impact on demand and supply structures after extracting each
#' sector \insertCite{miller_input-output_2009}{fio}.
#'
#' The total impact is calculated by the sum of the direct and indirect impacts.
#'
#' ## References
#' \insertCited{}
#' @return
#' Self (invisibly).
#' @examples
#' # data
#' intermediate_transactions <- matrix(c(1, 2, 3, 4, 5, 6, 7, 8, 9), 3, 3)
#' total_production <- matrix(c(100, 200, 300), 1, 3)
#' exports_data <- matrix(c(5, 10, 15), 3, 1)
#' holsehold_consumption_data <- matrix(c(20, 25, 30), 3, 1)
#' operating_income_data <- matrix(c(2, 5, 10), 1, 3)
#' taxes_data <- matrix(c(1, 2, 3), 1, 3)
#' # instantiate iom object
#' my_iom <- fio::iom$new(
#' "test",
#' intermediate_transactions,
#' total_production,
#' exports = exports_data,
#' household_consumption = holsehold_consumption_data,
#' operating_income = operating_income_data,
#' taxes = taxes_data
#' )
#' # update value-added matrix
#' my_iom$update_value_added_matrix()
#' # update final demand matrix
#' my_iom$update_final_demand_matrix()
#' # calculate the technical coefficients
#' my_iom$compute_tech_coeff()
#' # calculate the Leontief inverse
#' my_iom$compute_leontief_inverse()
#' # calculate allocation coefficients
#' my_iom$compute_allocation_coeff()
#' # calculate Ghosh inverse
#' my_iom$compute_ghosh_inverse()
#' # calculate hypothetical extraction
#' my_iom$compute_hypothetical_extraction()
#' # show results
#' my_iom$hypothetical_extraction
compute_hypothetical_extraction = function() {
# check if arguments are available
for (matrix_name in c(
"technical_coefficients_matrix",
"allocation_coefficients_matrix"
)) {
if (is.null(self[[matrix_name]])) {
cli::cli_h1("Error in {matrix_name}")
error(paste("You must compute the", matrix_name, "first. Run respective compute_*() method."))
}
}
for (matrix in c(
"final_demand_matrix",
"value_added_matrix"
)) {
if (is.null(self[[matrix_name]])) {
cli::cli_h1("Error in {matrix_name}")
error("You must compute the {matrix_name} first. Run respective update_*() method.")
}
}
# save row and column names
row_names <- rownames(self$technical_coefficients_matrix)
# compute backward extraction
extraction_backward <- compute_extraction_backward(
technical_coefficients_matrix = self$technical_coefficients_matrix,
final_demand_matrix = self$final_demand_matrix,
total_production = self$total_production
)
# compute forward extraction
extraction_forward <- compute_extraction_forward(
allocation_coeff = self$allocation_coefficients_matrix,
value_added_matrix = self$value_added_matrix,
total_production = self$total_production
)
# compute total extraction
extraction_total <- compute_extraction_total(
backward_linkage_matrix = extraction_backward,
forward_linkage_matrix = extraction_forward
)
# bind
hypothetical_extraction <- cbind(
extraction_backward,
extraction_forward,
extraction_total
)
# set row and column names
rownames(hypothetical_extraction) <- row_names
colnames(hypothetical_extraction) <- c(
"backward_absolute",
"backward_relative",
"forward_absolute",
"forward_relative",
"total_absolute",
"total_relative"
)
# store matrix
self$hypothetical_extraction <- hypothetical_extraction
invisible(self)
},
#' @description
#' Sets max number of threads used by fio and populate the `threads` field with the resulting `(integer)`.
#' @param max_threads (`integer`)\cr
#' Number of threads enabled for parallel computing. Defaults to 0, meaning all
#' threads available.
#' @details
#' Calling this function sets a global limit of threads to Rayon crate, affecting
#' all computations that runs in parallel by default.
#'
#' Default behavior of Rayon is to use all available threads (including logical). Setting to 1 will result in
#' single threaded (sequential) computations.
#'
#' Initialization of the global thread pool happens exactly once. Once started, the configuration cannot be changed
#' in the current session. If `$set_max_threads()` is called again in the same session, it'll result in an error.
#'
#' Methods that deals with linear algebra computations, like `$compute_leontief_inverse()` and
#' `$compute_ghosh_inverse()`, will try to use all available threads by default, so they also initializes global
#' thread pool. In order to choose a maximum number of threads other than default, `$set_max_threads()` must be
#' called before any computation, preferably right after `iom$new()`.
#' @return
#' This function does not return a value.
#' @examples
#' \dontrun{
#' intermediate_transactions <- matrix(c(1, 2, 3, 4, 5, 6, 7, 8, 9), 3, 3)
#' total_production <- matrix(c(100, 200, 300), 1, 3)
#' # instantiate iom object
#' my_iom <- fio::iom$new("test", intermediate_transactions, total_production)
#' # to run single threaded (sequential)
#' my_iom$set_max_threads(1L)
#' }
set_max_threads = function(max_threads) {
# assert type
if (!(is.integer(max_threads) && max_threads >= 0)) {
return(error("max_threads must be a positive integer."))
}
if (max_threads == 0) {
return(alert("0 means all available threads, which is default behavior. Nothing changed"))
}
return(set_max_threads(max_threads))
}
),
# private members
private = list(
iom_elements = function() {
c(
"intermediate_transactions",
"total_production",
"household_consumption",
"government_consumption",
"exports",
"final_demand_others",
"imports",
"taxes",
"wages",
"operating_income",
"value_added_others",
"occupation"
)
}
)
)
Try the fio package in your browser
Any scripts or data that you put into this service are public.
fio documentation built on April 11, 2025, 6:11 p.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
#' @title
#' R6 class for input-output matrix
#'
#' @description
#' R6 class for input-output matrix.
#'
#' @param id (`character`)\cr
#' Identifier for the input-output matrix.
#' @param intermediate_transactions (`matrix`)\cr
#' Intermediate transactions matrix.
#' @param total_production (`matrix`)\cr
#' Total production vector.
#' @param household_consumption (`matrix`)\cr
#' Household consumption vector.
#' @param government_consumption (`matrix`)\cr
#' Government consumption vector.
#' @param exports (`matrix`)\cr
#' Exports vector.
#' @param final_demand_others (`matrix`)\cr
#' Other vectors of final demand that doesn't have dedicated slots.
#' Setting column names is advised for better readability.
#' @param imports (`matrix`)\cr
#' Imports vector.
#' @param taxes (`matrix`)\cr
#' Taxes vector.
#' @param wages (`matrix`)\cr
#' Wages vector.
#' @param operating_income (`matrix`)\cr
#' Operating income vector.
#' @param value_added_others (`matrix`)\cr
#' Other vectors of value-added that doesn't have dedicated slots.
#' Setting row names is advised for better readability.
#' @param occupation (`matrix`)\cr
#' Occupation matrix.
#'
#' @return A new instance of the `iom` class.
#'
#' @examples
#' # data
#' intermediate_transactions <- matrix(c(1, 2, 3, 4, 5, 6, 7, 8, 9), 3, 3)
#' total_production <- matrix(c(100, 200, 300), 1, 3)
#' exports <- matrix(c(10, 20, 30), 3, 1)
#' households <- matrix(as.numeric(4:6), 3, 1)
#' imports <- matrix(c(5, 10, 15), 1, 3)
#' jobs <- matrix(c(10, 12, 15), 1, 3)
#' taxes <- matrix(c(2, 5, 10), 1, 3)
#' wages <- matrix(c(11, 12, 13), 1, 3)
#'
#' # a new iom instance can be created by passing just intermediate transactions and total production
#' my_iom <- iom$new(
#' "example_1",
#' intermediate_transactions,
#' total_production
#' )
#'
#' # or by passing optional arguments
#' my_iom <- iom$new(
#' "example_2",
#' intermediate_transactions,
#' total_production,
#' household_consumption = households,
#' exports = exports,
#' imports = imports,
#' taxes = taxes,
#' wages = wages,
#' occupation = jobs
#' )
#'
#' @importFrom Rdpack reprompt
#' @import R6
#' @export
# input-output matrix class
iom <- R6Class(
classname = "iom",
public = list(
#' @field id (`character`)\cr
#' Identifier of the new instance.
id = NULL,
#' @field intermediate_transactions (`matrix`)\cr
#' Intermediate transactions matrix.
intermediate_transactions = NULL,
#' @field total_production (`matrix`)\cr
#' Total production vector.
total_production = NULL,
#' @field household_consumption (`matrix`)\cr
#' Household consumption vector.
household_consumption = NULL,
#' @field government_consumption (`matrix`)\cr
#' Government consumption vector.
government_consumption = NULL,
#' @field exports (`matrix`)\cr
#' Exports vector.
exports = NULL,
#' @field final_demand_others (`matrix`)\cr
#' Other vectors of final demand that doesn't have dedicated slots.
final_demand_others = NULL,
#' @field final_demand_matrix (`matrix`)\cr
#' Aggregates final demand vectors into a matrix.
final_demand_matrix = NULL,
#' @field imports (`matrix`)\cr
#' Imports vector.
imports = NULL,
#' @field taxes (`matrix`)\cr
#' Taxes vector.
taxes = NULL,
#' @field wages (`matrix`)\cr
#' Wages vector.
wages = NULL,
#' @field operating_income (`matrix`)\cr
#' Operating income vector.
operating_income = NULL,
#' @field value_added_others (`matrix`)\cr
#' Other vectors of value-added that doesn't have dedicated slots.
value_added_others = NULL,
#' @field value_added_matrix (`matrix`)\cr
#' Aggregates value-added vectors into a matrix.
value_added_matrix = NULL,
#' @field occupation (`matrix`)\cr
#' Occupation vector.
occupation = NULL,
#' @field technical_coefficients_matrix (`matrix`)\cr
#' Technical coefficients matrix.
technical_coefficients_matrix = NULL,
#' @field leontief_inverse_matrix (`matrix`)\cr
#' Leontief inverse matrix.
leontief_inverse_matrix = NULL,
#' @field multiplier_output (`data.frame`)\cr
#' Output multiplier dataframe.
multiplier_output = NULL,
#' @field multiplier_employment (`data.frame`)\cr
#' Employment multiplier dataframe.
multiplier_employment = NULL,
#' @field multiplier_taxes (`data.frame`)\cr
#' Taxes multiplier dataframe.
multiplier_taxes = NULL,
#' @field multiplier_wages (`data.frame`)\cr
#' Wages multiplier dataframe.
multiplier_wages = NULL,
#' @field field_influence (`matrix`)\cr
#' Influence field matrix.
field_influence = NULL,
#' @field key_sectors (`data.frame`)\cr
#' Key sectors dataframe.
key_sectors = NULL,
#' @field allocation_coefficients_matrix (`matrix`)\cr
#' Allocation coefficients matrix.
allocation_coefficients_matrix = NULL,
#' @field ghosh_inverse_matrix (`matrix`)\cr
#' Ghosh inverse matrix.
ghosh_inverse_matrix = NULL,
#' @field hypothetical_extraction (`matrix`)\cr
#' Absolute and relative backward and forward differences in total output after a hypothetical extraction
hypothetical_extraction = NULL,
#' @description
#' Creates a new instance of this [R6][R6::R6Class] class.
initialize = function(id,
intermediate_transactions,
total_production,
household_consumption = NULL,
government_consumption = NULL,
exports = NULL,
final_demand_others = NULL,
imports = NULL,
taxes = NULL,
wages = NULL,
operating_income = NULL,
value_added_others = NULL,
occupation = NULL) {
### assertions ###
# check class
for (matrix in private$iom_elements()) {
if (!is.null(get_var(matrix)) && !is.matrix(get_var(matrix))) {
cli::cli_h1("Error in matrix class")
alert(paste("Try coerce", matrix, "to a matrix using as.matrix() function."))
error(paste(matrix, "must be a matrix."))
}
}
# check dimensions
if (nrow(intermediate_transactions) != ncol(intermediate_transactions)) {
cli::cli_h1("Error in matrix dimensions")
error("intermediate_transactions must be a square matrix")
}
for (matrix in c(
"household_consumption",
"government_consumption",
"exports",
"final_demand_others"
)) {
if (
!is.null(get_var(matrix))
&& nrow(get_var(matrix)) != nrow(intermediate_transactions)
) {
cli::cli_h1("Error in matrix dimensions")
error(
paste(
matrix,
"must have the same number of rows than `intermediate_transactions`,
which is",
nrow(intermediate_transactions),
"rows. But has",
nrow(get_var(matrix)),
"rows."
)
)
}
}
for (matrix in c(
"imports",
"taxes",
"wages",
"operating_income",
"value_added_others",
"occupation",
"total_production"
)) {
if (!is.null(get_var(matrix)) && ncol(get_var(matrix)) != ncol(intermediate_transactions)) {
cli::cli_h1("Error in matrix dimensions")
error(
paste(
matrix,
"must have the same number of columns than `intermediate_transactions`, which is",
ncol(intermediate_transactions),
"columns. But",
matrix,
"has",
ncol(get_var(matrix)),
"columns."
)
)
}
}
# check number format
for (matrix in private$iom_elements()) {
if (!is.null(get_var(matrix))) {
# Check if the matrix storage mode is not double
if (storage.mode(get_var(matrix)) != "double") {
cli::cli_h1("Error in matrix number format")
alert(
paste(
"Try coerce",
matrix,
"elements to double using as.numeric()."
)
)
error(paste(matrix, "elements must be of type double."))
}
}
}
# set data members
self$id <- id
self$intermediate_transactions <- intermediate_transactions
self$total_production <- total_production
self$household_consumption <- set_colnames(household_consumption)
self$government_consumption <- set_colnames(government_consumption)
self$exports <- set_colnames(exports)
self$final_demand_others <- final_demand_others
self$imports <- set_rownames(imports)
self$taxes <- set_rownames(taxes)
self$wages <- set_rownames(wages)
self$operating_income <- set_rownames(operating_income)
self$value_added_others <- value_added_others
self$occupation <- set_rownames(occupation)
},
#' @description
#' Adds a `matrix` to the `iom` object.
#' @param matrix_name (`character`)\cr
#' One of household_consumption, government_consumption, exports, final_demand_others,
#' imports, taxes, wages, operating income, value_added_others or occupation matrix to be added.
#' @param matrix (`matrix`)\cr
#' Matrix object to be added.
#' @return
#' Self (invisibly).
#' @examples
#' # data
#' intermediate_transactions <- matrix(c(1, 2, 3, 4, 5, 6, 7, 8, 9), 3, 3)
#' total_production <- matrix(c(100, 200, 300), 1, 3)
#' # instantiate iom object
#' my_iom <- iom$new("mock", intermediate_transactions, total_production)
#' # Create a dummy matrix
#' exports_data <- matrix(as.numeric(1:3), 3, 1)
#' # Add the matrix
#' my_iom$add("exports", exports_data)
add = function(matrix_name, matrix) {
# check arg
choices <- private$iom_elements()
if (!matrix_name %in% choices) {
cli::cli_h1("Error in matrix_name")
error(paste("matrix_name must be one of", paste(choices, collapse = ", ")))
}
# check class
if (!is.matrix(matrix)) {
cli::cli_h1("Error in matrix class")
alert(paste("Try coerce", matrix_name, "to a matrix using as.matrix() function."))
error(paste(matrix_name, "must be a matrix."))
}
# check dimensions
if (matrix_name %in% c("household_consumption", "government_consumption", "exports", "final_demand_others")) {
if (nrow(matrix) != nrow(self$intermediate_transactions)) {
cli::cli_h1("Error in matrix dimensions")
error(
paste(
matrix_name,
"must have the same number of rows than intermediate_transactions, which is",
nrow(self$intermediate_transactions),
"rows. But",
matrix_name,
"has",
nrow(matrix),
"rows."
)
)
}
} else {
if (ncol(matrix) != ncol(self$intermediate_transactions)) {
cli::cli_h1("Error in matrix dimensions")
error(
paste(
matrix_name,
"must have the same number of columns than intermediate_transactions, which is",
ncol(self$intermediate_transactions),
"columns. But", matrix_name, "has", ncol(matrix), "columns."
)
)
}
}
# import matrix
self[[matrix_name]] <- matrix
invisible(self)
},
#' @description
#' Removes a `matrix` from the `iom` object.
#' @param matrix_name (`character`)\cr
#' One of household_consumption, government_consumption, exports, final_demand_others,
#' imports, taxes, wages, operating_income, value_added_others or occupation matrix to be removed.
#' @return Self (invisibly).
#' @examples
#' # data
#' intermediate_transactions <- matrix(c(1, 2, 3, 4, 5, 6, 7, 8, 9), 3, 3)
#' total_production <- matrix(c(100, 200, 300), 1, 3)
#' exports_data <- matrix(as.numeric(1:3), 3, 1)
#' # instantiate iom object
#' my_iom <- iom$new("mock", intermediate_transactions, total_production, exports = exports_data)
#' # Remove the matrix
#' my_iom$remove("exports")
remove = function(matrix_name) {
# check arg
choices <- private$iom_elements()
if (!matrix_name %in% choices) {
cli::cli_h1("Error in matrix_name")
error(paste("matrix_name must be one of", paste(choices, collapse = ", ")))
}
# remove matrix
self[[matrix_name]] <- NULL
invisible(self)
},
#' @description
#' Aggregates final demand vectors into the `final_demand_matrix` field.
#' @details
#' Some methods, as `$compute_hypothetical_extraction()`, require the final demand and value-added vectors
#' to be aggregated into a matrix. This method does this aggregation, binding the vectors into
#' `$final_demand_matrix`.
#' @return
#' This functions doesn't returns a value.
#' @examples
#' # data
#' intermediate_transactions <- matrix(c(1, 2, 3, 4, 5, 6, 7, 8, 9), 3, 3)
#' total_production <- matrix(c(100, 200, 300), 1, 3)
#' exports_data <- matrix(c(10, 20, 30), 3, 1)
#' households <- matrix(as.numeric(4:6), 3, 1)
#' # instantiate iom object
#' my_iom <- iom$new(
#' "mock",
#' intermediate_transactions,
#' total_production,
#' exports = exports_data,
#' household_consumption = households
#' )
#' # aggregate all final demand vectors
#' my_iom$update_final_demand_matrix()
#' # check final demand matrix
#' my_iom$final_demand_matrix
update_final_demand_matrix = function() {
# bind final demand vectors
self$final_demand_matrix <- as.matrix(cbind(
self$household_consumption,
self$government_consumption,
self$exports,
self$final_demand_others
))
},
#' @description
#' Aggregates value-added vectors into the `value_added_matrix` field.
#' @details
#' Some methods, as `$compute_hypothetical_extraction()`, require the final demand and value-added vectors
#' to be aggregated into a matrix. This method does this aggregation, binding the vectors into
#' `$value_added_matrix`.
#' @return
#' This functions doesn't returns a value.
#' @examples
#' # data
#' intermediate_transactions <- matrix(c(1, 2, 3, 4, 5, 6, 7, 8, 9), 3, 3)
#' total_production <- matrix(c(100, 200, 300), 1, 3)
#' imports_data <- matrix(c(5, 10, 15), 1, 3)
#' taxes_data <- matrix(c(2, 5, 10), 1, 3)
#' # instantiate iom object
#' my_iom <- iom$new(
#' "mock",
#' intermediate_transactions,
#' total_production,
#' imports = imports_data,
#' taxes = taxes_data
#' )
#' # aggregate all value-added vectors
#' my_iom$update_value_added_matrix()
#' # check value-added matrix
#' my_iom$value_added_matrix
update_value_added_matrix = function() {
# bind value-added vectors
self$value_added_matrix <- as.matrix(rbind(
self$imports,
self$taxes,
self$wages,
self$operating_income,
self$value_added_others
))
},
#' @description
#' Computes the technical coefficients matrix and populate the `technical_coefficients_matrix` field with the
#' resulting `(matrix)`.
#' @details
#' It computes the technical coefficients matrix, a \eqn{n x n} matrix known as `A` matrix which is the column-wise
#' ratio of intermediate transactions to total production \insertCite{leontief_economia_1983}{fio}.
#'
#' ## References
#' \insertCited{}
#' @return
#' Self (invisibly).
#' @examples
#' intermediate_transactions <- matrix(c(1, 2, 3, 4, 5, 6, 7, 8, 9), 3, 3)
#' total_production <- matrix(c(100, 200, 300), 1, 3)
#' # instantiate iom object
#' my_iom <- iom$new("test", intermediate_transactions, total_production)
#' # Calculate the technical coefficients
#' my_iom$compute_tech_coeff()
#' # show the technical coefficients
#' my_iom$technical_coefficients_matrix
compute_tech_coeff = function() {
# save row and column names
row_names <- if (is.null(rownames(self$intermediate_transactions))) {
seq_len(nrow(self$intermediate_transactions))
} else {
rownames(self$intermediate_transactions)
}
col_names <- if (is.null(colnames(self$intermediate_transactions))) {
seq_len(ncol(self$intermediate_transactions))
} else {
colnames(self$intermediate_transactions)
}
# calculate technical coefficients matrix
technical_coefficients_matrix <- compute_tech_coeff(
intermediate_transactions = self$intermediate_transactions,
total_production = self$total_production
)
# set row and column names
rownames(technical_coefficients_matrix) <- row_names
colnames(technical_coefficients_matrix) <- col_names
# store matrix
self$technical_coefficients_matrix <- technical_coefficients_matrix
invisible(self)
},
#' @description
#' Computes the Leontief inverse matrix and populate the `leontief_inverse_matrix` field with the resulting
#' `(matrix)`.
#' @details
#' It computes the Leontief inverse matrix \insertCite{leontief_economia_1983}{fio}, which is the inverse of the
#' Leontief matrix, defined as:
#'
#' \deqn{L = I - A}
#'
#' where I is the identity matrix and A is the technical coefficients matrix.
#' The Leontief inverse matrix is calculated by solving the following equation:
#'
#' \deqn{L^{-1} = (I - A)^{-1}}
#'
#' Since the Leontief matrix is a square matrix and the subtraction of the technical coefficients matrix from the
#' identity matrix guarantees that the Leontief matrix is invertible, underlined Rust function uses LU decomposition
#' to solve the equation.
#'
#' ## References:
#' \insertCited{}
#' @return
#' Self (invisibly).
#' @examples
#' intermediate_transactions <- matrix(c(1, 2, 3, 4, 5, 6, 7, 8, 9), 3, 3)
#' total_production <- matrix(c(100, 200, 300), 1, 3)
#' # instantiate iom object
#' my_iom <- fio::iom$new("test", intermediate_transactions, total_production)
#' # calculate the technical coefficients
#' my_iom$compute_tech_coeff()
#' # calculate the Leontief inverse
#' my_iom$compute_leontief_inverse()
#' # show the Leontief inverse
#' my_iom$leontief_inverse_matrix
compute_leontief_inverse = function() {
# check if technical coefficients matrix is available
if (is.null(self$technical_coefficients_matrix)) {
cli::cli_h1("Error in technical_coefficients_matrix")
error("You must compute the technical coefficients matrix first. Run compute_tech_coeff() method.")
}
# save row and column names
row_names <- rownames(self$technical_coefficients_matrix)
col_names <- colnames(self$technical_coefficients_matrix)
# computes leontief inverse matrix
leontief_inverse_matrix <- compute_leontief_inverse(
tech_coeff = self$technical_coefficients_matrix
)
# set row and column names
rownames(leontief_inverse_matrix) <- row_names
colnames(leontief_inverse_matrix) <- col_names
# store matrix
self$leontief_inverse_matrix <- leontief_inverse_matrix
invisible(self)
},
#' @description
#' Computes the output multiplier and populate the `multiplier_output` field with the resulting `(data.frame)`.
#' @details
#' An output multiplier for sector *j* is defined as the total value of production in all sectors of the economy
#' that is necessary in order to satisfy a monetary unit (e.g., a dollar) worth of final demand for sector *j*'s
#' output \insertCite{miller_input-output_2009}{fio}.
#'
#' This method computes the simple output multiplier, defined as the column sums of the Leontief inverse matrix,
#' the direct and indirect output multipliers, which are the column sums of the technical
#' coefficients matrix and the difference between total and direct output multipliers, respectively
#' \insertCite{vale_alise_2020}{fio}.
#'
#' ## References
#' \insertCited{}
#' @return
#' Self (invisibly).
#' @examples
#' # data
#' intermediate_transactions <- matrix(c(1, 2, 3, 4, 5, 6, 7, 8, 9), 3, 3)
#' total_production <- matrix(c(100, 200, 300), 1, 3)
#' # instantiate iom object
#' my_iom <- fio::iom$new("test", intermediate_transactions, total_production)
#' # calculate the technical coefficients
#' my_iom$compute_tech_coeff()
#' # calculate the Leontief inverse
#' my_iom$compute_leontief_inverse()
#' # calculate the output multiplier
#' my_iom$compute_multiplier_output()
#' # show the output multiplier
#' my_iom$multiplier_output
compute_multiplier_output = function() {
# check if leontief inverse matrix is available
if (is.null(self$leontief_inverse_matrix)) {
cli::cli_h1("Error in leontief_inverse_matrix")
error("You must compute the leontief inverse matrix first. Run compute_leontief_inverse() method.")
}
# save column names
col_names <- colnames(self$leontief_inverse_matrix)
# compute output multiplier vector
multiplier_output_simple <- compute_multiplier_output(
leontief_inverse_matrix = self$leontief_inverse_matrix
)
# compute direct output multiplier vector
multiplier_output_direct <- compute_multiplier_output_direct(
technical_coefficients_matrix = self$technical_coefficients_matrix
)
# compute indirect output multiplier vector
multiplier_output_indirect <- compute_multiplier_output_indirect(
technical_coefficients_matrix = self$technical_coefficients_matrix,
leontief_inverse_matrix = self$leontief_inverse_matrix
)
multiplier_output <- data.frame(
sector = col_names,
multiplier_simple = multiplier_output_simple,
multiplier_direct = multiplier_output_direct,
multiplier_indirect = multiplier_output_indirect
)
# store vector
self$multiplier_output <- multiplier_output
invisible(self)
},
#' @description
#' Computes the employment multiplier and populate the `multiplier_employment` field with the resulting
#' `(data.frame)`.
#' @details
#' The employment multiplier for sector *j* relates the jobs created in each sector in response to a
#' initial exogenous shock \insertCite{miller_input-output_2009}{fio}.
#'
#' Current implementation follows \insertCite{vale_alise_2020}{fio}.
#'
#' ## References
#' \insertCited{}
#' @return
#' Self (invisibly).
#' @examples
#' # data
#' intermediate_transactions <- matrix(c(1, 2, 3, 4, 5, 6, 7, 8, 9), 3, 3)
#' total_production <- matrix(c(100, 200, 300), 1, 3)
#' jobs_data <- matrix(c(10, 12, 15), 1, 3)
#' # instantiate iom object
#' my_iom <- fio::iom$new("test", intermediate_transactions, total_production, occupation = jobs_data)
#' # calculate the technical coefficients
#' my_iom$compute_tech_coeff()
#' # calculate the Leontief inverse
#' my_iom$compute_leontief_inverse()
#' # calculate the employment multiplier
#' my_iom$compute_multiplier_employment()
#' # show the employment multiplier
#' my_iom$multiplier_employment
compute_multiplier_employment = function() {
# check if leontief inverse matrix is available
if (is.null(self$leontief_inverse_matrix)) {
cli::cli_h1("Error in leontief_inverse_matrix")
error("You must compute the leontief inverse matrix first. Run compute_leontief_inverse() method.")
}
# save column names
col_names <- colnames(self$leontief_inverse_matrix)
# compute employment requirements
employment_requirements <- compute_requirements_value_added(
value_added_element = self$occupation,
total_production = self$total_production
)
# compute employment multiplier vector
multiplier_employment_simple <- compute_multiplier_value_added(
value_added_requirements = employment_requirements,
leontief_inverse_matrix = self$leontief_inverse_matrix
)
# compute indirect employment multiplier
multiplier_employment_indirect <- compute_multiplier_value_added_indirect(
value_added_element = self$occupation,
total_production = self$total_production,
leontief_inverse_matrix = self$leontief_inverse_matrix
)
multiplier_employment <- data.frame(
sector = col_names,
multiplier_simple = multiplier_employment_simple,
multiplier_direct = employment_requirements,
multiplier_indirect = multiplier_employment_indirect
)
# store vector
self$multiplier_employment <- multiplier_employment
invisible(self)
},
#' @description
#' Computes the wages multiplier dataframe and populate the `multiplier_wages` field with the resulting
#' `(data.frame)`.
#' @details
#' The wages multiplier for sector *j* relates increases in wages for each
#' sector in response to a initial exogenous shock
#' \insertCite{miller_input-output_2009}{fio}.
#'
#' Current implementation follows \insertCite{vale_alise_2020}{fio}.
#'
#' ## References
#' \insertCited{}
#' @return
#' Self (invisibly).
#' @examples
#' # data
#' intermediate_transactions <- matrix(c(1, 2, 3, 4, 5, 6, 7, 8, 9), 3, 3)
#' total_production <- matrix(c(100, 200, 300), 1, 3)
#' wages_data <- matrix(c(10, 12, 15), 1, 3)
#' # instantiate iom object
#' my_iom <- fio::iom$new("test", intermediate_transactions, total_production, wages = wages_data)
#' # calculate the technical coefficients
#' my_iom$compute_tech_coeff()
#' # calculate the Leontief inverse
#' my_iom$compute_leontief_inverse()
#' # calculate the wages multiplier
#' my_iom$compute_multiplier_wages()
#' # show the wages multiplier
#' my_iom$multiplier_wages
compute_multiplier_wages = function() {
# check if leontief inverse matrix is available
if (is.null(self$leontief_inverse_matrix)) {
cli::cli_h1("Error in leontief_inverse_matrix")
error("You must compute the leontief inverse matrix first. Run compute_leontief_inverse() method.")
}
# save column names
col_names <- colnames(self$leontief_inverse_matrix)
# compute wages requirements
wages_requirements <- compute_requirements_value_added(
value_added_element = self$wages,
total_production = self$total_production
)
# compute wages multiplier vector
multiplier_wages_simple <- compute_multiplier_value_added(
value_added_requirements = wages_requirements,
leontief_inverse_matrix = self$leontief_inverse_matrix
)
# compute indirect wages multiplier
multiplier_wages_indirect <- compute_multiplier_value_added_indirect(
value_added_element = self$wages,
total_production = self$total_production,
leontief_inverse_matrix = self$leontief_inverse_matrix
)
multiplier_wages <- data.frame(
sector = col_names,
multiplier_simple = multiplier_wages_simple,
multiplier_direct = wages_requirements,
multiplier_indirect = multiplier_wages_indirect
)
# store vector
self$multiplier_wages <- multiplier_wages
invisible(self)
},
#' @description
#' Computes the taxes multiplier and populate the `multiplier_taxes` field with
#' the resulting `(data.frame)`.
#' @details
#' The taxes multiplier for sector *j* relates the increases on tax revenue from
#' each sector in response to a initial exogenous shock
#' \insertCite{miller_input-output_2009}{fio}.
#'
#' Current implementation follows \insertCite{vale_alise_2020}{fio}.
#'
#' ## References
#' \insertCited{}
#' @return
#' Self (invisibly).
#' @examples
#' # data
#' intermediate_transactions <- matrix(c(1, 2, 3, 4, 5, 6, 7, 8, 9), 3, 3)
#' total_production <- matrix(c(100, 200, 300), 1, 3)
#' tax_data <- matrix(c(10, 12, 15), 1, 3)
#' # instantiate iom object
#' my_iom <- fio::iom$new("test", intermediate_transactions, total_production, taxes = tax_data)
#' # calculate the technical coefficients
#' my_iom$compute_tech_coeff()
#' # calculate the Leontief inverse
#' my_iom$compute_leontief_inverse()
#' # calculate the tax multiplier
#' my_iom$compute_multiplier_taxes()
#' # show the taxes multiplier
#' my_iom$multiplier_taxes
compute_multiplier_taxes = function() {
# check if leontief inverse matrix is available
if (is.null(self$leontief_inverse_matrix)) {
cli::cli_h1("Error in leontief_inverse_matrix")
error("You must compute the leontief inverse matrix first. Run compute_leontief_inverse() method.")
}
# save column names
col_names <- colnames(self$leontief_inverse_matrix)
# compute taxes requirements
taxes_requirements <- compute_requirements_value_added(
value_added_element = self$taxes,
total_production = self$total_production
)
# compute taxes multiplier vector
multiplier_taxes_simple <- compute_multiplier_value_added(
value_added_requirements = taxes_requirements,
leontief_inverse_matrix = self$leontief_inverse_matrix
)
# compute indirect taxes multiplier
multiplier_taxes_indirect <- compute_multiplier_value_added_indirect(
value_added_element = self$taxes,
total_production = self$total_production,
leontief_inverse_matrix = self$leontief_inverse_matrix
)
multiplier_taxes <- data.frame(
sector = col_names,
multiplier_simple = multiplier_taxes_simple,
multiplier_direct = taxes_requirements,
multiplier_indirect = multiplier_taxes_indirect
)
# store vector
self$multiplier_taxes <- multiplier_taxes
invisible(self)
},
#' @description
#' Computes the field of influence for all sectors and populate the
#' `field_influence` field with the resulting `(matrix)`.
#' @details
#' The field of influence shows how changes in direct coefficients are
#' distributed throughout the entire economic system, allowing for the
#' determination of which relationships between sectors are most important
#' within the production process.
#'
#' It determines which sectors have the greatest influence over others,
#' specifically, which coefficients, when altered, would have the greatest
#' impact on the system as a whole \insertCite{vale_alise_2020}{fio}.
#'
#' ## References
#' \insertCited{}
#' @param epsilon (`numeric`)\cr
#' Epsilon value. A technical change in the input-output matrix, caused by a variation of size `epsilon` into each
#' element of technical coefficients matrix.
#' @return
#' Self (invisibly).
#' @examples
#' # data
#' intermediate_transactions <- matrix(c(1, 2, 3, 4, 5, 6, 7, 8, 9), 3, 3)
#' total_production <- matrix(c(100, 200, 300), 1, 3)
#' # instantiate iom object
#' my_iom <- fio::iom$new("test", intermediate_transactions, total_production)
#' # calculate the technical coefficients
#' my_iom$compute_tech_coeff()
#' # calculate the Leontief inverse
#' my_iom$compute_leontief_inverse()
#' # calculate field of influence
#' my_iom$compute_field_influence(epsilon = 0.01)
#' # show the field of influence
#' my_iom$field_influence
compute_field_influence = function(epsilon) {
# check if epsilon was set
if (missing(epsilon)) {
cli::cli_h1("Error in epsilon")
error("You must set the epsilon value.")
}
# check if leontief inverse matrix is available
if (is.null(self$leontief_inverse_matrix)) {
cli::cli_h1("Error in leontief_inverse_matrix")
error("You must compute the leontief inverse matrix first. Run compute_leontief_inverse() method.")
}
# save row and column names
row_names <- rownames(self$technical_coefficients_matrix)
col_names <- colnames(self$technical_coefficients_matrix)
# compute influence field matrix
field_influence <- compute_field_influence(
tech_coeff = self$technical_coefficients_matrix,
leontief_inverse_matrix = self$leontief_inverse_matrix,
epsilon = epsilon
)
# set row and column names
rownames(field_influence) <- row_names
colnames(field_influence) <- col_names
# store matrix
self$field_influence <- field_influence
invisible(self)
},
#' @description
#' Computes the key sectors dataframe, based on it's power and sensitivity of dispersion,
#' and populate the `key_sectors` field with the resulting `(data.frame)`.
#' @details
#' Increased production from a sector *j* means that the sector *j* will need to
#' purchase more goods from other sectors. At the same time, it means that more goods from sector *j* will be
#' available for other sectors to purchase. Sectors that are above average in the demand sense (stronger backward
#' linkage) have power of dispersion indices greater than 1. Sectors that are above average in the supply sense
#' (stronger forward linkage) have sensitivity of dispersion indices greater than 1
#' \insertCite{miller_input-output_2009}{fio}.
#'
#' As both power and sensitivity of dispersion are related to average values on the economy, coefficients of
#' variation are also calculated for both indices. The lesser the coefficient of variation, greater the number of
#' sectors on the demand or supply structure of that sector \insertCite{vale_alise_2020}{fio}.
#'
#' ## References
#' \insertCited{}
#' @return
#' Self (invisibly).
#' @examples
#' # data
#' intermediate_transactions <- matrix(c(1, 2, 3, 4, 5, 6, 7, 8, 9), 3, 3)
#' total_production <- matrix(c(100, 200, 300), 1, 3)
#' # instantiate iom object
#' my_iom <- fio::iom$new("test", intermediate_transactions, total_production)
#' # calculate the technical coefficients
#' my_iom$compute_tech_coeff()
#' # calculate the Leontief inverse
#' my_iom$compute_leontief_inverse()
#' # calculate key sectors
#' my_iom$compute_key_sectors()
#' # show the key sectors
#' my_iom$key_sectors
compute_key_sectors = function() {
# check if leontief inverse matrix is available
if (is.null(self$leontief_inverse_matrix)) {
cli::cli_h1("Error in leontief_inverse_matrix")
error("You must compute the leontief inverse matrix first. Run compute_leontief_inverse() method.")
}
# power of dispersion
power_dispersion <- compute_power_dispersion(
leontief_inverse_matrix = self$leontief_inverse_matrix
)
# sensitivity of dispersion
sensitivity_dispersion <- compute_sensitivity_dispersion(
leontief_inverse_matrix = self$leontief_inverse_matrix
)
# power of dispersion coefficients of variation
power_dispersion_cv <- compute_power_dispersion_cv(
leontief_inverse_matrix = self$leontief_inverse_matrix
)
# sensitivity of dispersion coefficients of variation
sensitivity_dispersion_cv <- compute_sensitivity_dispersion_cv(
leontief_inverse_matrix = self$leontief_inverse_matrix
)
# compute key sectors dataframe
key_sectors <- data.frame(
sector = rownames(self$leontief_inverse_matrix),
power_dispersion = power_dispersion,
sensitivity_dispersion = sensitivity_dispersion,
power_dispersion_cv = power_dispersion_cv,
sensitivity_dispersion_cv = sensitivity_dispersion_cv
) |>
within({
key_sectors <- ifelse(sensitivity_dispersion <= 1 & power_dispersion <= 1, "Non-Key Sector", "")
key_sectors <- ifelse(sensitivity_dispersion > 1 & power_dispersion > 1, "Key Sector", key_sectors)
key_sectors <- ifelse(sensitivity_dispersion > 1 & power_dispersion <= 1, "Strong Forward Linkage", key_sectors) # nolint
key_sectors <- ifelse(sensitivity_dispersion <= 1 & power_dispersion > 1, "Strong Backward Linkage", key_sectors) # nolint
})
# store dataframe
self$key_sectors <- key_sectors
invisible(self)
},
#' @description
#' Computes the allocation coefficients matrix and populate the `allocation_coefficients_matrix` field with the
#' resulting `(matrix)`.
#' @details
#' It computes the allocation coefficients matrix, a \eqn{n x n} matrix known as `B` matrix which is the row-wise
#' ratio of intermediate transactions to total production \insertCite{miller_input-output_2009}{fio}.
#'
#' ## References
#' \insertCited{}
#' @return
#' Self (invisibly).
#' @examples
#' intermediate_transactions <- matrix(c(1, 2, 3, 4, 5, 6, 7, 8, 9), 3, 3)
#' total_production <- matrix(c(100, 200, 300), 1, 3)
#' # instantiate iom object
#' my_iom <- fio::iom$new("test", intermediate_transactions, total_production)
#' # Calculate the allocation coefficients
#' my_iom$compute_allocation_coeff()
#' # show the allocation coefficients
#' my_iom$allocation_coefficients_matrix
compute_allocation_coeff = function() {
# save row and column names
row_names <- if (is.null(rownames(self$intermediate_transactions))) {
seq_len(nrow(self$intermediate_transactions))
} else {
rownames(self$intermediate_transactions)
}
col_names <- if (is.null(colnames(self$intermediate_transactions))) {
seq_len(ncol(self$intermediate_transactions))
} else {
colnames(self$intermediate_transactions)
}
# compute allocation coefficients matrix
allocation_coefficients_matrix <- compute_allocation_coeff(
intermediate_transactions = self$intermediate_transactions,
total_production = self$total_production
)
# set row and column names
rownames(allocation_coefficients_matrix) <- row_names
colnames(allocation_coefficients_matrix) <- col_names
# store matrix
self$allocation_coefficients_matrix <- allocation_coefficients_matrix
invisible(self)
},
#' @description
#' Computes the Ghosh inverse matrix and populate the `ghosh_inverse_matrix` field with the resulting `(matrix)`.
#' @details
#' It computes the Ghosh inverse matrix \insertCite{miller_input-output_2009}{fio}, defined as:
#' \deqn{G = (I - B)^{-1}}
#' where I is the identity matrix and B is the allocation coefficients matrix.
#'
#' ## References
#' \insertCited{}
#' @return
#' Self (invisibly).
#' @examples
#' intermediate_transactions <- matrix(c(1, 2, 3, 4, 5, 6, 7, 8, 9), 3, 3)
#' total_production <- matrix(c(100, 200, 300), 1, 3)
#' # instantiate iom object
#' my_iom <- fio::iom$new("test", intermediate_transactions, total_production)
#' # Calculate the allocation coefficients
#' my_iom$compute_allocation_coeff()
#' # Calculate the Ghosh inverse
#' my_iom$compute_ghosh_inverse()
#' # show the Ghosh inverse
#' my_iom$ghosh_inverse_matrix
compute_ghosh_inverse = function() {
# check if allocation coefficients matrix is available
if (is.null(self$allocation_coefficients_matrix)) {
cli::cli_h1("Error in allocation_coefficients_matrix")
error("You must compute the allocation coefficients matrix first. Run compute_allocation_coeff() method.")
}
# save row and column names
row_names <- rownames(self$allocation_coefficients_matrix)
col_names <- colnames(self$allocation_coefficients_matrix)
# compute ghosh inverse matrix
ghosh_inverse_matrix <- compute_ghosh_inverse(
allocation_coeff = self$allocation_coefficients_matrix
)
# set row and column names
rownames(ghosh_inverse_matrix) <- row_names
colnames(ghosh_inverse_matrix) <- col_names
# store matrix
self$ghosh_inverse_matrix <- ghosh_inverse_matrix
invisible(self)
},
#' @description
#' Computes total impact after extracting a each sector and populate the `hypothetical_extraction` field with the
#' resulting `(data.frame)`.
#' @details
#' Computes impact on demand and supply structures after extracting each
#' sector \insertCite{miller_input-output_2009}{fio}.
#'
#' The total impact is calculated by the sum of the direct and indirect impacts.
#'
#' ## References
#' \insertCited{}
#' @return
#' Self (invisibly).
#' @examples
#' # data
#' intermediate_transactions <- matrix(c(1, 2, 3, 4, 5, 6, 7, 8, 9), 3, 3)
#' total_production <- matrix(c(100, 200, 300), 1, 3)
#' exports_data <- matrix(c(5, 10, 15), 3, 1)
#' holsehold_consumption_data <- matrix(c(20, 25, 30), 3, 1)
#' operating_income_data <- matrix(c(2, 5, 10), 1, 3)
#' taxes_data <- matrix(c(1, 2, 3), 1, 3)
#' # instantiate iom object
#' my_iom <- fio::iom$new(
#' "test",
#' intermediate_transactions,
#' total_production,
#' exports = exports_data,
#' household_consumption = holsehold_consumption_data,
#' operating_income = operating_income_data,
#' taxes = taxes_data
#' )
#' # update value-added matrix
#' my_iom$update_value_added_matrix()
#' # update final demand matrix
#' my_iom$update_final_demand_matrix()
#' # calculate the technical coefficients
#' my_iom$compute_tech_coeff()
#' # calculate the Leontief inverse
#' my_iom$compute_leontief_inverse()
#' # calculate allocation coefficients
#' my_iom$compute_allocation_coeff()
#' # calculate Ghosh inverse
#' my_iom$compute_ghosh_inverse()
#' # calculate hypothetical extraction
#' my_iom$compute_hypothetical_extraction()
#' # show results
#' my_iom$hypothetical_extraction
compute_hypothetical_extraction = function() {
# check if arguments are available
for (matrix_name in c(
"technical_coefficients_matrix",
"allocation_coefficients_matrix"
)) {
if (is.null(self[[matrix_name]])) {
cli::cli_h1("Error in {matrix_name}")
error(paste("You must compute the", matrix_name, "first. Run respective compute_*() method."))
}
}
for (matrix in c(
"final_demand_matrix",
"value_added_matrix"
)) {
if (is.null(self[[matrix_name]])) {
cli::cli_h1("Error in {matrix_name}")
error("You must compute the {matrix_name} first. Run respective update_*() method.")
}
}
# save row and column names
row_names <- rownames(self$technical_coefficients_matrix)
# compute backward extraction
extraction_backward <- compute_extraction_backward(
technical_coefficients_matrix = self$technical_coefficients_matrix,
final_demand_matrix = self$final_demand_matrix,
total_production = self$total_production
)
# compute forward extraction
extraction_forward <- compute_extraction_forward(
allocation_coeff = self$allocation_coefficients_matrix,
value_added_matrix = self$value_added_matrix,
total_production = self$total_production
)
# compute total extraction
extraction_total <- compute_extraction_total(
backward_linkage_matrix = extraction_backward,
forward_linkage_matrix = extraction_forward
)
# bind
hypothetical_extraction <- cbind(
extraction_backward,
extraction_forward,
extraction_total
)
# set row and column names
rownames(hypothetical_extraction) <- row_names
colnames(hypothetical_extraction) <- c(
"backward_absolute",
"backward_relative",
"forward_absolute",
"forward_relative",
"total_absolute",
"total_relative"
)
# store matrix
self$hypothetical_extraction <- hypothetical_extraction
invisible(self)
},
#' @description
#' Sets max number of threads used by fio and populate the `threads` field with the resulting `(integer)`.
#' @param max_threads (`integer`)\cr
#' Number of threads enabled for parallel computing. Defaults to 0, meaning all
#' threads available.
#' @details
#' Calling this function sets a global limit of threads to Rayon crate, affecting
#' all computations that runs in parallel by default.
#'
#' Default behavior of Rayon is to use all available threads (including logical). Setting to 1 will result in
#' single threaded (sequential) computations.
#'
#' Initialization of the global thread pool happens exactly once. Once started, the configuration cannot be changed
#' in the current session. If `$set_max_threads()` is called again in the same session, it'll result in an error.
#'
#' Methods that deals with linear algebra computations, like `$compute_leontief_inverse()` and
#' `$compute_ghosh_inverse()`, will try to use all available threads by default, so they also initializes global
#' thread pool. In order to choose a maximum number of threads other than default, `$set_max_threads()` must be
#' called before any computation, preferably right after `iom$new()`.
#' @return
#' This function does not return a value.
#' @examples
#' \dontrun{
#' intermediate_transactions <- matrix(c(1, 2, 3, 4, 5, 6, 7, 8, 9), 3, 3)
#' total_production <- matrix(c(100, 200, 300), 1, 3)
#' # instantiate iom object
#' my_iom <- fio::iom$new("test", intermediate_transactions, total_production)
#' # to run single threaded (sequential)
#' my_iom$set_max_threads(1L)
#' }
set_max_threads = function(max_threads) {
# assert type
if (!(is.integer(max_threads) && max_threads >= 0)) {
return(error("max_threads must be a positive integer."))
}
if (max_threads == 0) {
return(alert("0 means all available threads, which is default behavior. Nothing changed"))
}
return(set_max_threads(max_threads))
}
),
# private members
private = list(
iom_elements = function() {
c(
"intermediate_transactions",
"total_production",
"household_consumption",
"government_consumption",
"exports",
"final_demand_others",
"imports",
"taxes",
"wages",
"operating_income",
"value_added_others",
"occupation"
)
}
)
)
Try the fio package in your browser
Any scripts or data that you put into this service are public.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.