TableFactor Fixes

[varunagrawal]

Found some issues in the TableFactor while debugging an incorrect assignment on conversion from DiscreteConditional to TableFactor. The fixes are listed below.

1. Fix the equals method so it also checks the keys are the same. Updated a unit test to show this.
2. Updated printing so the vertical bars are neatly aligned.
3. Check if the size of the provided table matches the maximum cardinality of the keys. Else we get a weird result with collisions like below:

TableFactor:
 f[ (0,2), (1,3), ]
(0, 0)(1, 0) | 0.166667   | 0  // Repeat
(0, 0)(1, 1) | 0.277778   | 1
(0, 0)(1, 2) | 0.3        | 2  // Repeat
(0, 0)(1, 0) | 0.333333   | 3  // Repeat
(0, 1)(1, 1) | 0.333333   | 4  // Repeat
(0, 1)(1, 2) | 0.333333   | 5
(0, 1)(1, 0) | 0.5        | 6
(0, 1)(1, 1) | 0.388889   | 7  // Repeat
(0, 0)(1, 2) | 0.366667   | 8  // Repeat
number of nnzs: 9

I am also working on improving the conversion from DecisionTreeFactor to TableFactor in a way that should be faster and more memory-efficient than the current scheme. I'll probably push that to the next PR.

[dellaert]

Stronger for it!