A short list of the main features of TraDemgEn is given below sorted in different categories. Many more features and functions exist and for these we refer to the reference documentation.

Demand generation

The demand can be generated thanks to two relatively advanced pieces of algorithm, both following a sequential principle. That is, the events (booking requests) are generated one after the other, sequentially, rather than being generated all at once at the beginning of the process (e.g., a simulation).

The two sequential methods are:

• 'Intuitive' method. The booking period is sliced in intervals, where the arrival rate of events (booking requests) is known for each of those intervals, say . The inter-arrival process then follows an exponential law. That is, the final number of booking requests follows a Non homogeneous Poisson distribution. With that method, the variance of that distribution is therefore equal to the mean.
• 'Advanced' method. The process uses order statistics in order to mimic the behaviour of uniform distributions projected onto the know arrival pattern of events. With that method, the final number of booking requests is first drawn, following any probability distribution (e.g., normal, Gamma, Beta or even Weibull law) with any required standard deviation. Then, each booking request is drawn in sequence:
  • according to a mere uniform distribution,
  • and projected onto the known booking arrival pattern.

Other features

• CSV input file parsing
• Memory handling