We will augment the original x2
dataframe with this new information by merging back by category, and then by sorting the dataframe by the rank of the coefficient. This will allow us to use this as a proxy for trend:
x2x <- x2 %>% left_join(xx4, by = "cat") %>% arrange(coef.rank, cat) # exclude some columns so as to fit on one page head(x2x[, c(-2, -3, -4, -8)]) > Source: local data frame [6 x 7] > > cat Year.1 Total.People Total Not.Covered.Pct > (fctr) (int) (dbl) (dbl) (dbl) > 1 MALE 18 to 24 YEARS 2012 15142.04 11091.86 0.2674787 > 2 MALE 18 to 24 YEARS 2011 15159.87 11028.75 0.2725034 > 3 MALE 18 to 24 YEARS 2010 14986.02 10646.88 0.2895460 > 4 MALE 18 to 24 YEARS 2010 14837.14 10109.82 0.3186139 > 5 MALE 18 to 24 YEARS 2008 14508.04 10021.66 0.3092339 > 6 MALE 18 to 24 YEARS 2007...