I write code that runs in high throughput production environments expected to handle a variety of input from multiple users. In this environment, I often need to join (with dplyr) two multi-column dataframes with join columns of mismatched types which produces this error. "Can't join on ... because of incompatible types"
The production code is expected to handle input csvs that are 150+ columns by 0-1000 rows, with 12-20 join columns. I use read.table for speed, and because the content of the 150 columns can vary, allow type to be inferred by data content.
It would be great to have a general-purpose function that matches join column types programmatically, making no assumptions about column name or location.