functions to calclute adjacency matrix

as_tbl_graph.corr.test <- function(x, ...) {

  as_tbl_graph(as.igraph(x, ...))

}

#' @noRd

adjacency_matrix <- function(x, ...) {

  UseMethod("adjacency_matrix")

}

#' @method adjacency_matrix cor_md_tbl

adjacency_matrix.cor_md_tbl <- function(x, ...) {

  if (isTRUE(attr(x, "grouped"))) {

    stop("Cannot convert a grouped md_tbl to a adjacency_matrix.", call. = FALSE)

  }

  rnm <- row_names(x)

  cnm <- col_names(x)

  x <- dplyr::filter(x, ...)

  df_to_matrix(x, "r", row_id = ".rownames", col_id = ".colnames",

               row_names = rnm, col_names = cnm, missing = 0)

}

#' @method adjacency_matrix default

adjacency_matrix.default <- function(x, ...) {

  adjacency_matrix(as_md_tbl(x), ...)

}

#' @noRd

df_to_matrix <- function(x,

                         value,

                         row_id = NULL,

                         col_id = NULL,

                         row_names = NULL,

                         col_names = NULL,

                         missing = NA) {

  row_id <- row_id %||% names(x)[1]

  col_id <- col_id %||% names(x)[2]

  rnm <- row_names %||% unique(x[[row_id]])

  cnm <- col_names %||% unique(x[[col_id]])

  ID <- paste(rep(rnm, length(cnm)), rep(cnm, each = length(rnm)), sep = "--")

  vv <- rep(missing, length(ID))

  ii <- match(paste(x[[row_id]], x[[col_id]], sep = "--"), ID)

  vv[ii] <- x[[value]]

  matrix(vv, nrow = length(rnm), ncol = length(cnm),

         dimnames = list(rnm, cnm))

}