Standarize nomenclature

[javicid]

Standarize nomenclature in the source code and documentation according to:

1. dislib.Array instances are referred to as ds-array.
2. NumPy array instances are referred to as ndarray.
3. x and y should be used for ds-arrays representing samples and target values.
4. x_np and y_np should be used for ndarrays representing samples and target values.
5. A NumPy array or csr_matrix that is a part of a ds-array should be named block.
6. When iterating ds-arrays horizontally and vertically, hblock and vblock should be used to refer to sets of blocks.
7. Tasks that receive a set of blocks as input parameter should name this parameter blocks.
8. In the documentation, variable x should be described as 'Training samples'.
9. In the documentation, variable y should be described as 'Target values'.
10. Optional arguments should be documented in the formula: "int, optional (default=0)"
11. Input ds-arrays should be documented with the formula: "ds-array, shape=(n_samples, n_features)"
12. The name of functions that are tasks should start with _.
13. The name of files and functions that are not supposed to be accessed by users should start with _.
14. Estimators need to be implemented in a base.py file in a separate sub-folder inside the appropriate submodule. Additional files can be included in the same subfolder named with a leading _.
15. Other typical variable names:

• number of something = n_something
• max_iter
• arity
• tol = tolerance criteria
• random_state
• verbose
• check_convergence (whether to check for convergence)