longnet: What the Package Does (One Line, Title Case)

# library(dplyr)
# library(igraph)
# library(lubridate)
# library(rlang)
#
# ## testing only
# # library(readr)
# #
# # data <- read_csv("./Data/Full_Data.csv")
# #
# # node_var = "IA.ID"
# # edge_var = "Member.ID"
# # start_date_var = "Start.Date"
# # end_date_var = "End.Date"
# # first_date = "1970-01-01"
# # last_date = NA
# # date_orders = "ymd"
# # node_labels = "IA.Name"
# # edge_labels = "Full.Name"
# # # node_remove = NULL
# # # edge_remove = NULL
#
# #
# #
# make_graph_df <- function(data,
#                        node_var, edge_var,
#                        start_date_var = NULL, end_date_var = NULL,
#                        first_date = NA, last_date = NA,
#                        date_orders = "ymd",
#                        node_labels = NA, edge_labels = NA,
#                        node_remove = NULL, edge_remove = NULL){
#
#   # start_date_var and end_date_var are column names
#
#   # first_date and last_date should not both be NA.
#   ## If they are, give a warning and use the whole dataset.
#
#   # node_var and edge_var specify the variable names in the dataset to use
#   ## eventually I should tidyeval this
#
#   # node_labels and edge_labels optionally specify columns with better names
#   ## REMOVED just attach metas
#
#   # node_remove and edge_remove are lists with
#   ## [[1]] name of dataset column to reference,
#   ## [[2]] vector of values to omit
#
#
#   #### Filter dataset to relevant date range, and optional omissions ####
#
#   drop <- rep(FALSE, nrow(data))
#
#   if (!is.null(node_remove)) {
#
#     drop <- drop | data[[ node_remove[[1]] ]] %in% node_remove[[2]]
#
#   }
#
#   if (!is.null(edge_remove)) {
#
#     drop <- drop | data[[ edge_remove[[1]] ]] %in% edge_remove[[2]]
#
#   }
#
#   ## need to throw some errors here if weird stuff is supplied.
#   ## being lazy for now because it's my data I won't be stupid (hopefully)
#   ## start date var is required; end date will be considered start date if not supplied
#
#   if (!is.null(start_date_var)) {
#
#     data[[start_date_var]] <- lubridate::parse_date_time(data[[start_date_var]], date_orders)
#
#     if (!is.null(end_date_var)) {
#
#       data[[end_date_var]] <- lubridate::parse_date_time(data[[end_date_var]], date_orders)
#
#     } else {
#
#       end_date_var <- "end_date"
#       data[[end_date_var]] <- data[[start_date_var]]
#
#     }
#
#     if (is.na(first_date)) {
#
#       first_date = min(data[[start_date_var]])
#
#     } else {
#
#       first_date = lubridate::parse_date_time(first_date, orders = date_orders)
#
#     }
#
#     if (is.na(last_date)) {
#
#       last_date = max(data[[end_date_var]])
#
#     } else {
#
#       last_date = lubridate::parse_date_time(last_date, orders = date_orders)
#
#     }
#
#     drop <- drop | data[[end_date_var]] < first_date | data[[start_date_var]] > last_date
#
#   }
#
#   # Aaand, drop 'em
#   data <- data[!drop, ]
#
#   #### Time to make a graph! ####
#
#
#   # Make graph from counting co-membership
#   my_graph <- data[,c(node_var, edge_var)] %>%
#     dplyr::group_by(.data[[node_var]]) %>%
#     table() %>%
#     as.matrix %>%
#     tcrossprod() %>%
#     graph_from_adjacency_matrix(weighted = TRUE) %>%
#     get.data.frame()
#
#
#   #### Add edge info ####
#
#   my_graph <- my_graph %>%
#     mutate(
#       edge_members = map2_chr(to, from, ~find_edge_members(data, node_var, edge_var, .x, .y))
#       )
#
#   # Attach names and meta-info to data frame
#     return(my_graph)
#   }