Improve auto categorisations

[RayStick]

Currently the code automatically categorizes certain variables if they contain these strings, see the line in the code that starts

  # auto categorise (full string and partial string matches)

This minimises the number of variables that the user has to sort through, and automatically sorts the obvious ones.

1. Are there other common variable names that pop up in SAIL datasets that we could auto sort?

• LSOA_CD lower super output area
• PRAC_CD_E anonymized general practice identifier
• BIRTH_MARRIAGE_INDCD - this is being picked up as AGE because it finds 'AGE' this is a :bug: