Removed import of the unicode_literal for all files. This will be added back only when h5py is fixed.

Changed how we define custom dtypes to work when import unicode_literals from __future__

Importing all necessary functions to ensure python 3 compatibility for all notebooks and package files.
Note - Extensive testing is pending

Fixed swapping of rows and columns when setting up the datasets.

Incremented version number

Translator now names the dataset correctly

Translators now store the noise floor as a compound dataset

Updated to use new methods from Translator baseclass
Uses new calc_chunks function

Moved the functions and fixed imports/calls in io

Words are split with an underscore now

Cleaned up more print statements and CamelCased the class names, etc.