HARplus (version 1.1.2)

save_har: Save Data to GEMPACK HAR Format

Description

Writes one or more data matrices or arrays into a GEMPACK-compatible HAR file format, automatically generating associated 1C set headers.

Usage

save_har(
  data_list,
  file_path,
  dimensions,
  value_cols = NULL,
  long_desc = NULL,
  coefficients = NULL,
  export_sets = TRUE,
  lowercase = TRUE,
  dim_order = NULL,
  dim_rename = NULL
)

Value

Invisibly returns a list containing export metadata, including file path and header summary.

Arguments

data_list

A named list of data frames or arrays (names up to 4 characters).

file_path

Path to the output HAR file.

dimensions

A named list specifying dimension columns for each header.

• HAR dimension names follow these column names; rename columns before saving to change them.

value_cols

A named vector of value column names. Defaults to "Value".

long_desc

A named vector of long descriptions (optional).

coefficients

A named vector of coefficient names (optional).

export_sets

Logical; if TRUE, exports dimension sets as 1C headers. Default is TRUE.

lowercase

Logical; if TRUE, converts elements to lowercase. Default is TRUE.

dim_order

Optional dimension ordering specification. Can be:

• NULL (default): alphabetical A–Z ordering

• a data frame with columns matching dimension names

• a named list specifying order for each dimension

• a character string giving the path to an Excel or CSV file with order definitions

dim_rename

Optional named list to rename dimensions in HAR output.

• Names are original column names, values are desired HAR dimension names

• Allows duplicate dimension names in HAR (e.g., COMMxREGxREG from COMMxSREGxDREG)

• Example: list(RTMS = c(COMM = "COMM", SREG = "REG", DREG = "REG"))

Author

Pattawee Puangchit

Details

• Supports 1CFULL (string sets) and REFULL (real arrays) headers

• 2IFULL (integer) and RESPSE (sparse) types are not implemented

• Up to seven dimensions and ~2 million elements per chunk

• No practical file-size limit or header count restriction

See Also

load_harx, load_sl4x

Examples

# Example 1: Save a single matrix
har_path <- system.file("extdata", "TAR10-WEL.har", package = "HARplus")
har_data <- load_harx(har_path)
welfare_data <- get_data_by_var("A", har_data)
welfare_data_A <- welfare_data[["har_data"]][["A"]]

save_har(
  data_list   = list(WELF = welfare_data),
  file_path   = file.path(tempdir(), "output_single.har"),
  dimensions  = list(WELF = c("REG", "COLUMN")),
  value_cols  = c(WELF = "Value"),
  long_desc   = c(WELF = "Welfare Decomposition"),
  export_sets = TRUE,
  lowercase   = FALSE
)

# Example 2: Save multiple matrices
welfare_data <- get_data_by_var(c("A", "A1"), har_data)
welfare_data <- welfare_data[["har_data"]]

save_har(
  data_list   = list(WELF = welfare_data[["A"]],
                     DECOM = welfare_data[["A1"]]),
  file_path   = file.path(tempdir(), "output_multi.har"),
  dimensions  = list(WELF = c("REG", "COLUMN"),
                     DECOM = c("REG", "ALLOCEFF")),
  value_cols  = list(WELF = "Value",
                     DECOM = "Value"),
  long_desc   = list(WELF = "Welfare Decomposition",
                     DECOM = "Allocative efficiency effect"),
  export_sets = TRUE,
  lowercase   = FALSE
)

# Example 3: Apply mapping file for sorting
mapping <- list(
  REG    = c("RestofWorld", "MENA", "EU_28"),
  COLUMN = c("alloc_A1", "tot_E1")
)

save_har(
  data_list   = list(
    WELF  = welfare_data[["A"]],
    DECOM = welfare_data[["A1"]]
  ),
  file_path   = file.path(tempdir(), "output_sorted.har"),
  dimensions  = list(
    WELF  = c("REG", "COLUMN"),
    DECOM = c("REG", "ALLOCEFF")
  ),
  value_cols  = list(
    WELF  = "Value",
    DECOM = "Value"
  ),
  long_desc   = list(
    WELF  = "Welfare Decomposition",
    DECOM = "Allocative efficiency effect"
  ),
  export_sets = TRUE,
  dim_order   = mapping,
  lowercase   = FALSE
)

# Example 4: Rename duplicated dimension names  
# to export as a single structure (e.g., COMMxREGxREG)
har_path <- system.file("extdata", "TAR10-WEL.har", package = "HARplus")
har_data <- load_harx(har_path)
welfare_data <- get_data_by_var("C17", har_data)
welfare_data <- welfare_data[["har_data"]][["C17"]] 

mapping <- list(
  COMM = c("TransComm", "MeatLstk"),
  REG  = c("SSA", "RestofWorld", "SEAsia")  
)

# Export HAR
save_har(
  data_list   = list(TEST = welfare_data),
  file_path   = file.path(tempdir(), "output_single.har"),
  dimensions  = list(TEST = c("COMM", "REG", "REG.1")),
  value_cols  = list(TEST = "Value"),
  long_desc   = list(TEST = "Shock RTMS"),
  dim_rename  = list(TEST = c(COMM = "COMM", SREG = "REG", DREG = "REG.1")),
  export_sets = TRUE,
  dim_order   = mapping,
  lowercase   = FALSE
)