Add visualization

[zStupan]

• Added method for visualizing association rules with hill slopes, as presented in this article.
• Added a conditional statement to the rule constructor, allowing rules to be constructed without requiring transactions to be passed in,

I will add examples and more documentation on visualization in the future. And probably an improved interface.

Basic usage:

import matplotlib.pyplot as plt
from niaarm.visualize import tdf

# load dataset...
# get rules...

fig, ax = tdf(rule, transactions)
plt.show()

[firefly-cpp]

Wonderful! May I recommend changing the title of the visualization scheme from tdf to hill slopes (e.g. hill_slopes())?