The GaussianNormalizer should accept distribution names that are consistent with scipy

[npatki]

Problem Description

The GaussianNormalizer accepts a distribution parameter, where you can provide a string name of the univariate distribution.

The acceptable strings are: 'gaussian', 'gamma', 'beta', 'student_t', 'truncated_gaussian' and 'gaussian_kde'

These names are inconsistent with both of the following:

• The underlying scipy library where the univariates come from and
• The SDV library which uses the same names as scipy.

Expected behavior

Change the mappings to the following:

• 'norm' should be used for a Gaussian distribution
• 'truncnorm' should be used for a truncated Gaussian Distribution
• 't' should be used for Student T

Additional context

In order to keep backwards compatibility, we can still accept the older names. If a user inputs them, internally use the newer names and show a Future Warning.

# example: user inputs 'gaussian' instead of 'norm'
my_transformer = GaussianNormalizer(distribution='gaussian')

This ultimately uses the correct distribution but it shows a future warning.

FutureWarning: Future versions of RDT will not support 'gaussian' as an option. Please use 'norm' instead.