Speed optimisations

[jpsrobertson]

Right now this package reads in data extremely slowly. This is a master issue for exploring ways to further optimise the function for speed.

Ideas:

• address the for loop #5
• reading only specific columns in rather than the entire table #3
• review the data frame binding #6
• support for multi threading #4
• implementation of some sort of in memory database #7