Convert Coutinuous Variable to Categorical

Suppose x is a continuous variable, we want to convert it to categorical. For example, we want to assign 1 to top 30%, 2 to middle 40%, and 3 to last 30%.

Stata

We can use xtile or egen cut in stata.

set obs 1000
gen x =  rnormal()

xtile xc = x,  nquantiles(10)
tabulate xc

// equal frequency grouping intervals
egen xc2 = cut(x), group(10)
tabulate xc2

R

cut{base} divides the range of x into intervals and codes the values in x according to which interval they fall. The leftmost interval corresponds to level one, the next leftmost to level two and so on.

cut(x, 
    quantile(x, prob=seq(0,1,1/5)), 
    labels=c(1:5), 
    include.lowest=TRUE)

Matlab

ordinal is what you want, ref.
See also categorical.

p = 0:.25:1;
x = randn(1000,1);
breaks = quantile(x, p);
xc = ordinal(x, {'Q1','Q2','Q3','Q4'}, [], breaks);
getlevels(xc)