Real data, such as from the structural brain network, is often assumed to consist of strong, likely real, edges and weak edges that could possibly be noise. In our project, we will add noise to a model network (navy) that has been thresholded to a specified edge density ρT (rose). For any edges not in the rose network, we will add random weights to these edges and thus create an network of added noise (gold). We can then measure the affect of noise on the structure of the combined network by expanding the weighted network into a sequence of binary networks and computing our favorite graph metric. If we create this sequence of graphs (right) by adding edges according to decreasing edge weights, we can see first the structure of the model edges, then after ρT we will see how randomly added edges will affect the structure as perceived by the graph descriptor.