R/simnet_df_perturb.R

In CommKern: Network-Based Communities and Kernel Machine Methods

#' Simulated network data frame
#'
#' Description of the simulated network data frame function.
#'
#' This is an ancillary function that creates a list of data frames, of which
#' each data frame describes the community assignment for each node in the
#' network. These data frames are used as a starting point for the edge weights
#' to be added between nodes (see \code{\link{group_network_perturb}} and
#' \code{\link{get_weights}} for more information).
#'
#' @param n_nodes the number of nodes in each simulated network (will be the
#' same across all networks)
#' @param n_comm the number of communities to be simulated in each network (will
#' be the same across all networks)
#' @param n_nets the number of networks to simulate
#' @param perturb_prop the proportion of network nodes to randomly alter their
#' community assignment within each network
#'
#' @return a list of network data frames containing nodes, their community
#' assignment, and node dyads
#'
simnet_df_perturb <- function(n_nodes, n_comm, n_nets, perturb_prop) {

    # Create a master data frame that has IDs for all nodes and assigns random
    # communities to nodes
    master_node_df <- 
      data.frame(Node = paste0("node_", formatC(1:n_nodes, format = "d", width = nchar(n_nodes), flag = "0")),
                 Comm = sample(x = paste0("comm_", letters[1:n_comm]), size = n_nodes, replace = TRUE))

    net_list <- list()
    perturb_num <- ceiling(n_nodes * perturb_prop)

    for (i in 1:n_nets) {
        rows <- sample(1:n_nodes, perturb_num)
        new_comms <- sample(x = paste0("comm_", letters[1:n_comm]), size = perturb_num,
            replace = T)

        ind_node_df <- master_node_df

        for (j in 1:length(rows)) {
            ind_node_df[rows[j], ][2] <- new_comms[j]
        }

        # Create a data frame with all possible combinations of nodes except
        # 'self-relationships'
        ind_dataframe <-
          expand.grid(Node_A = ind_node_df$Node,
                      Node_B = ind_node_df$Node,
                      stringsAsFactors = FALSE)

        ind_dataframe <- merge(x = ind_dataframe,
                               y = stats::setNames(ind_node_df, c("Node_B", "Node_B_Comm")),
                               all.x = TRUE,
                               all.y = FALSE,
                               by = "Node_B")
        ind_dataframe <- merge(x = ind_dataframe,
                               y = stats::setNames(ind_node_df, c("Node_A", "Node_A_Comm")),
                               all.x = TRUE,
                               all.y = FALSE,
                               by = "Node_A")
        ind_dataframe <- subset(ind_dataframe, ind_dataframe$Node_A != ind_dataframe$Node_B)

        # Limit the data frame to one row per dyad for an undirected network
        ind_dataframe <- ind_dataframe[order(ind_dataframe$Node_A, ind_dataframe$Node_B), ]
        ind_dataframe$dyad <- ifelse(ind_dataframe$Node_A < ind_dataframe$Node_B,
                                     paste(ind_dataframe$Node_A, ind_dataframe$Node_B, sep = "_"),
                                     paste(ind_dataframe$Node_B, ind_dataframe$Node_A, sep = "_"))
        ind_dataframe <- ind_dataframe[!duplicated(ind_dataframe$dyad), ]


        net_list[[i]] <- ind_dataframe
    }

    return(net_list)
}