Avoid significant logic inside of for loops

[cmatKhan]

OK -- not quite accurate, since using map or apply is just a for loop under the hood.

But, the point is that for loops that have significant logic in them are difficult to debug.

What is better is to take a page from the book of functional programming and write the
logic which executes on any given iteration of the loop into a function (or a number of functions)
and then apply that function(s) over the input list.

So, plot_metrics, rather than this:

 for (i in 1:length(metric_dfs_by_net[[1]])) {
    metric_dfs[[i]] <- do.call(cbind, lapply(unname(metric_dfs_by_net), "[[", i))
  }

  # assemble y-axis labels
  to_plot <- c(sum, pe
  ... # it goes on for many lines ....

write the logic inside of that loop into a function

[cmatKhan]

Same here

[cmatKhan]

I'm not sure what exactly is happening here -- there are some nested for loops and the if statement is long. will need to run the code.

But, rather than using a print statement, in general using a logger is cleaner. Gives both you and the user a lot more control over how, where and when the message is printed. I like futile.logger in R.