Basic data massaging-Open to better ideas

[AxisMeetsWorld]

I am trying to input data on various employment statistics from Universities around the nation. Problem is that one data source might write a university with an abbreviation, while another might not (IE "Oregon State" vs "Oregon St."). When I import new data in from a new source I use the setdiff() function to find the school representations that aren't found in the names I have decided on for them. I then keep a dataframe (call it convert dataframe) of representations I find in the new set and what I would like them to be converted to for consistency. I have a user-defined function that allows me to type in what I had, and the converted representation. This updates the convert dataframe with the new values. I finally have one more function that I can use to run through a new dataframe that has these discrepancies and look at the convert dataframe so that the newly imported information is consistent with university representations already stored. I am just wondering if this is the easiest way, or if there might be something else I should try. I will be at the R user-group tomorrow so I can show if necessary. Thanks!

[sckott]

thanks @AxisMeetsWorld - will add this to the agenda for tomorrow

[laderast]

forcats has the fct_collapse() function which may be really helpful for mapping these.

[AxisMeetsWorld]

Here is a link to the code https://docs.google.com/document/d/1scaOtvlRbBbNUp9jrEyigqREW6gezC1R3R5ark6GUm8/edit

[sckott]

@AxisMeetsWorld code isn't public, and you make it public?

[AxisMeetsWorld]

Should be public now