NA/NaN/Inf error when fitting LQMM

Question

I'm trying to replicate some results that were published in "Linear Quantile mixed models", Geraci M. and Bottai M. (2013). The data can be downloaded from: http://www.hsph.harvard.edu/fitzmaur/ala/tlc.txt

I am using linear quantile mixed model, with natural cubic splines (df=3, break points 0, 4, 6) and random effects. I have created the splines using

t1.t2 <-ns(Lead.r$week, df=3, knots=4)
t1.t2
t1 <- t1.t2[1:400, 2]
t1

The quantile mixed model is

.

I can manage to get an output when I model the splines and treatment (interaction and main effects), as given by:

fit.int2 <-lqmm(lead ~ t1.t2*group, random= ~1, group=id, 
            data=Lead.r, tau=c(0.25, 0.5, 0.75))

However, once I try to put in the T1 and treatment term (main effects and interaction), as below, I get an error.

fit.int2 <-lqmm(lead ~ t1.t2*group+t1*group, random= ~1, group=id, 
            data=Lead.r, tau=c(0.25, 0.5, 0.75))
Error: NA/NaN/Inf in foreign function call (arg 1)

I'm pretty new to R and quantile mixed models so I'm not sure if this is an error on my behalf, being coding, or lack of knowledge!