I'm running the quantile regression and trying to plot the coefficients and confidence intervals of four independent variables (yeaedu, sex, union, unimem).
The data a .dta file, but I used as_factor to convert data before running the regression.
bri0 = read.dta("britearn.dta")
bri0.1 = as_factor(bri0,only_labelled = TRUE)
model.bri.q = rq(lrehrwge ~ yeaedu + sex + union + unimem
- seniorit + yeactexp + race + mastat
- jbsect + region,
data = bri0, tau = seq(0.1, 0.9, by=0.1),
na.action = na.exclude)
My data file:
My errors are:
.... Solution may be nonunique
1: In summary.rq(xi, U = U, ...) : 37 non-positive fis 2: In summary.rq(xi, U = U, ...) : 9 non-positive fis 3: In summary.rq(xi, U = U, ...) : 2 non-positive fis 4: In summary.rq(xi, U = U, ...) : 9 non-positive fis 5: In summary.rq(xi, U = U, ...) : 5 non-positive fis 6: In summary.rq(xi, U = U, ...) : 7 non-positive is
So, my questions are:
- How did I get these errors?
- How can I plot the effects of only four variables yeaedu + sex + union + unimem on the dependent variable (most of the rest are categorical variables, I don't want to plot them)?
I'm a beginner user of R, so I hope that you can help me to resolve my problem well.
Thank you so much!