DKQS does not choose tau automatically if the user does not specify it

[a-torgovitsky]

This conflicts with the behavior discussed in the README.md whereby if the user does not pass tau then we automatically compute the largest feasible one and use that.

devtools::load_all()
source("../lpinfer/example/dgp_mixedlogit.R")

dgp <- mixedlogit_dgp()
set.seed(1)
data <- mixedlogit_draw(dgp, n = 2000)

lpm <- lpmodel(A.obs = mixedlogit_Aobs(dgp),
               beta.obs = function(d) mixedlogit_betaobs(d, dgp),
               A.shp = rep(1, nrow(dgp$vdist)),
               beta.shp = 1,
               A.tgt = mixedlogit_Atgt_dfelast(dgp, w2eval = 1, eeval = -1))
dkqs(data = data, lpm, .5)

gives

 Error: The object 'tau' has to be a nonnegative number. 

[conroylau]

Done! I have coded the part to choose tau when the user does not specify it, but there was a bug in the part that checks the arguments so it results in the error. It has been updated and the above code now gives the following:

 p-value: 1                                                                               
 tau used: 1

Thanks!