create pathological data set

[zkamvar]

• NAs in all columns
• missing ids in linelist
• missing ids in contacts
• clusters of known sizes
• degrees of known sizes
• unconnected cases
• node and edge attributes of all data types (numeric/character/date/POSIXt/factor/logical)
• duplicate IDs

[zkamvar]

A good place to start is the data set Finlay used for testing clusters:

  ## make clusters of different sizes
  cs_size <- c(5, 10, 17, 3, 5)
  contacts <- data.frame(from = 1:(sum(cs_size)-1), to = 2:sum(cs_size))
  contacts <- contacts[-head(cumsum(cs_size), -1),]
  linelist <- data.frame(id = 1:sum(cs_size))
  
  x <- make_epicontacts(linelist, contacts)  

I think we can build off of this