So I am working with some Census API data. Sadly there are a number of NAs in some of the columns. I am replacing the NAs with the column means, but I figured I would get more accurate information if I split the data by county first. Here is where my problem lies; I am unable to merge them back to a single dataframe correctly. I know that unsplit doesn't work for lists of dataframes so instead as per other posts I am using
do.call("rbind", hdi_tract$county)
instead. It seems to have worked but now I am getting more NA values than what I started with before splitting the data. Why is this the case?