We present a study of complex flows by implementing numerical simulations of the two-dimensional defocusing Gross-Pitaevskii equation with a stochastic band-limited randomly walked forcing function. Building on previous results, we aim to extend the quantitative understanding of the features of complicated flows through the use of Dynamic-Mode Decomposition (DMD). By clustering the primary DMD modes we are able to see the natural grouping of the modes and thereby use clustering as a means for dimensional reduction. Moreover, we use temporal averaging, statistical analysis, and machine learning to study the features of our complex flows. By implementing these data-driven methodologies, we are able to extract coherent states, such as vortices, present in our condensate.