I have done some preliminary analyses on the data. The results are [here](https://googledrive.com/host/0B4cyIPgV_VxraU5pTVFjNkFtYmM).
The same directory also contains the source as both an ipython notebook and a plain python script.
The main takeaway is in the scatter plot and the last cross correlation plot.
The dependence between the anomaly and the link strength looks weak and the dependence of the crosscorrelation on the lag is erratic.