I'm studying in the university in first year of Master aMIV which is about modelisation,biostatistics and bioinformatics. I've worked on this project during my training course which lasted from February to May 2006. The aim of my training course was to perform DynW. The most important option that I've added concerns the possibility of random variation of parameters or initial values that can represent biodiversity or sampling errors. I've also proposed two new options, the first one which permits to choose between two scales (the initial scale and the logarithmic one) and the second one which enables people to export their graphs into different formats (jpeg, png, eps, ps) which is very useful to include graphs in reports and especially those which are made in LateX. Finally, the last main point of my training course consisted in helping users by correcting the possible errors and informing them about the typing mistakes they can have made. DynW is made to be used by most people and in particular biologists,so hence it had to be pratical and simple of use.