R/filter_historical_mif.R

In pik-piam/mrremind: MadRat REMIND Input Data Package

#' Filter improperly aggregated regional data from historical.mif
#'
#' Load a `historical.mif` file, remove designated data, write back to file.
#'
#' For all combinations of scenario, model, and variable (and all years), either
#' all regions listed in `exclude_regions` in the `filter_table` are removed, or
#' all regions _except_ those listed in `include_regions` in the `filter_table`
#' are removed.
#' Defining both `include_regions` and `exclude_regions` on the same row of
#' `filter_table` will throw an error.
#' Scenarios, models, and variables are matched precisely, not partially or via
#' regular expressions.
#'
#' @md
#' @param path Path to `historical.mif` (or any mif for that matter), defaults
#'   to `historical.mif` in the currently configured
#'   `getConfig('outputfolder')`.
#' @param filter_table Data frame with columns `scenario`, `model`, `variable`,
#'   `include_regions`, and `exclude_regions`.  Defaults to the built-in
#'   `./inst/extdata/historical_mif_filter_table.csv` of the `mrremind` package.
#'
#' @importFrom madrat getConfig
#' @importFrom magclass getItems mbind read.report write.report
#' @importFrom quitte cartesian

#' @export
filter_historical_mif <- function(path = NULL, filter_table = NULL) {
    if (is.null(path))
        path <- 'historical.mif'

    if (is.null(filter_table))
        filter_table <- read.csv(
            file = system.file('extdata', 'historical_mif_filter_table.csv',
                               package = 'mrremind', mustWork = TRUE),
            colClasses = 'character')

    # combine scenario, model, and variable columns to string for indexing a
    # magpie object
    scenario_model_variable <- sapply(
        seq_len(nrow(filter_table)),
        function(i) {
            Reduce(f = cartesian,
                   x = filter_table[i, c('scenario', 'model', 'variable')])
        })

    # check for conflicting include/exclude definitions
    include_and_exclude <- ( '' != filter_table[['include_regions']]
                           & '' != filter_table[['exclude_regions']])

    if (any(include_and_exclude))
        stop('Both include and exclude regions defined for:',
             paste(c('', scenario_model_variable[include_and_exclude]),
                   collapse = '\n'))

    h <- read.report(file = path, as.list = FALSE)

    # process each row of the filter table
    for (i in seq_len(nrow(filter_table))) {
        if ('' != filter_table[[i, 'include_regions']]) {
          excludedRegions <- setdiff(getItems(h, dim = 'region'),
                                     filter_table[[i, 'include_regions']])
        } else {
          excludedRegions <- filter_table[[i, 'exclude_regions']]
        }

        h_include <- h[, , scenario_model_variable[i]]
        h_include[excludedRegions, , ] <- NA

        h <- mbind(h[, , scenario_model_variable[i], invert = TRUE], h_include)
    }

    write.report(h, file = path)
}