Documentation convention with consistency

[NiklasPfister]

Conventions:

• Parameters should be described with full sentences (i.e., starting with a capital letter and ending with a dot).
• only mention default values when explaining the exact behavior (as defaults are added automatically this avoids mentioning them twice)
• For parameters that have string values, mention possible values using " "
• For parameters that have the same behavior as sklearn, use the same description so it is easier for users to know where to expect the same behavior.
• Index matrices and vectors using square brackets, i.e., X[i, j] for a matrix and v[j] for a matrix. Try to use i for rows and j for columns whenever possible.
• Sample size should be denoted by n and number of features by p. In particular X is a (n x p)-matrix.
• X contains the "feature values" and Y the "response values"
• All input and output types should be defined.

[NiklasPfister]

Add a page "Contributing" to documentation.

[svbrodersen]

Add internal documentation convention. What functions should have what commens?

[svbrodersen]

Update features to be X and otucomes to be Y throughout the codebase.