Initialize Cramér-von Mises test

[RobbeBohy]

This PR introduces the first component of the new 3_validation workflow, as discussed in #27, and includes:

• Initialization of the workflow.
• A validation script that checks whether the sampled trajectories are consistent with the theoretical autocorrelation function (ACF), by comparing the distribution of $x(t + \Delta_t) - x(t)$ against the predicted distribution using the Cramér–von Mises test.

[tovrstra]

Just some minor suggestions below and a few general ones here:

• I would remove the _list suffixes of the variables names. The naming conventions I've used so far is somewhat fortranic, but works well in Python too: If a variable or an array represents a count of some sort, then prefix it with n (short for number of). If a variable is a list or array, add an s suffix (plural). Most people inuitively use an s suffix for both, which makes sense in a natural language, but does not work as a naming convention. Adding _list or _array instead of s is also common, but goes a bit against Python's duck typing idea.
• I've tested the code checked the plots. It all looks clear. It may be helpful to add the "p-value of the p-values somewhere on or above the bottom panel. (There may be better names for this thing.)

[tovrstra]

Minor suggestion: It would be more Pythonic to combine dts and pool into a single dictionary.

[RobbeBohy]

Good idea, I refactored this in 902d772.

[RobbeBohy]

Thanks for the comments. I have added the p-value in a box on the bottom panel like this: