Proposal
118 Term Project Progress Graph Problem Solver : Link Analysis Using Bryan's code and networkx manual for reference, I was able to add options for the user to pick a Link Analysis problem and upload a csv graph. They choose between PageRank or HITS problem solver and depending on which one the user clicks, they are directed to another menu. If they picked the pagerank subproblem, they can view the pagerank results through either numpy/scipy interfaces to evaluate the eigenvector calculations through different libraries. It returns a pop-up printing the pagerank dictionary. The menu also lets them view the google numpy transition matrix used in pageranking the graph. I was able to use the pagerank library to label nodes with their pagerank values on the graph, but the values don't align with the nodes. If they picked the hits subproblem, they can view the hits results through either numpy/scipy interfaces to evaluate and read both the hubs and authority vecotrs. There is also the option to read the authority or hits matrix in a pop-up. This coming week I will keep working on the code and hopefully draw out the graph results from the dictionary outputs from the functions.
118 Term Project Progress Graph Problem Solver : Link Analysis During week 6 discussion, I asked Bryan about the graph results I got from the dictionary results. Renaming the nodes to the corresponding dictionary value wasn't working with the networkx graphing functions. This coming week, I will try to implement a legend to go next to the graph drawing to show the pagerank and hits dictionary outputs. I was also told that flask cannot be used to develop the webapp. My goal for later is to research and work on using TomCat Apache with python.
