Visualisations for Schapiro's experiments

[Anirk4IR]

• Visualisations.py generates four plots by reading the results of a particular experiment. Two for the initial response (20 cycles aka study steps) and two for the settled response (80 cycles). On of the two plots is a heatmap representing Pearson correlation coefficient between the outputs of AHA when inputs are single characters. The value corresponds to the average across the networks (different seeds). The other plot is a line chart for the probability of generating an item at each epoch.
• Sequence generator was changed to include the core sequence for statistical learning, which is AB, CD, EF, GH, BC, BE, BG, DA, DE, DG, FA, FD, FG, HA, HC, HE; for 8 characters.
• Pairs structure changed to calculate the Pearson correlations and save the results.