There are two example data sets, each with 1000 observations, that were generated in R using the datGen function in the lrt.r file. These 
correspond to a null model case of no change point (tsDatH_0.txt) and an alternative with a change point at time 500 (tsDatH_1.txt). The 
seed for both cases was set at 123.

The C++ code is compiled using the make file lrt.mk. This should produce output that looks something like

my_directory_name$ make -f lrt.mk
g++ -c lrtDr.cpp 
g++ -c lrt.cpp
g++ -c matrix.cpp
g++ -c vector.cpp
g++ -o lrt lrtDr.o lrt.o matrix.o vector.o 

This end product is the executable lrt that is used with 3 command line arguments: the sample size, the order of the AR model and the 
file name for the data. For example, using this with the two data sets from R gives 

my_directory_name$ ./lrt 1000 2 tsDatH_0.txt
The test statistic is 5.76064
The estimated change point is 19

my_directory_name$ ./lrt 1000 2 tsDatH_1.txt
The test statistic is 129.244
The estimated change point is 508

These results can be compared with the output from the LRT function coded in R.
Specifically, we have

> set.seed(123)
> x <- datGen(1000)
> result <- LRT(x, 2)
> result[[2]]#estimated change point
[1] 19
> result[[1]]#test statistic
[1] 5.760639
> set.seed(123)
> x <- datGen(1000, .5, .1, .1)
> result <- LRT(x, 2)
> result[[2]]#estimated change point
[1] 508
> result[[1]]#test statistic
[1] 129.2440