Using the previous Pagerank to initialize a new calculation and Erdos graph generation

Simulation

We propose an algorithm that simulates the evolution of the web graph from one month to the next, based on an Erdos graph generation algorithm (update_erdos.py)

An implementation of Pagerank (implementation_PageRank.py)

Comparison

A tool to compare the converging speed of Pagerank initialized with the vector $$x$$ defined by $$\left(x_i\right)^t_{i\ \in {1,..,N}}$$ with $$x_i = \frac1N \ \forall \ i \ and \ \forall \ t$$ with Pagerank initialized with the previous month's result.

Matrix format

You can find matrixs to test the code here https://sparse.tamu.edu/SNAP/soc-Slashdot0902

Requirements

Type : square binary

Format : .mtx