No, there is no jump here. It is just use a binomial distribution to generate biased random walk. Ordinary Weiner process is based on normal distribution which as we all know can be replicated with throwing a fair coin (0.5,0.5) for infinite times. Here, this code is just to generate a biased distribution with a biased coin (p,q). On each interval "inc" it throws nx=inc/dt which is 100 in this case. 100 trials binomial distribution is roughly enough to approximate a continuous distribution.