Move visualisation and assign query into their own programs

[johnlees]

Mostly moving chunks of main() around, and some changes to the command line options. Once this is finished and merged rewriting the docs (#74) will be a priority

• Change --fit-model to take options gmm, dbscan, refine, threshold or use
• Remove --full-db from assign_query
• Get rid of easy run
• Ensure return_dict = True in printExternalClusters()
• Check for zero sequences to remove in reference_pick.py L84-85
• Check that query with a single sequence works (may get zero length self dist matrix)
• Check assign with an older model
• Check visualise with an older model
• Check assign with a lineage model (all ranks output)
• Check visualise with a lineage model
• Visualise needs to work with assign query output. It will require update db, and should still output query/ref in the table
• Visualise with assigned model + lineage mode (all ranks output and Visualisation with assign query + lineage fit #120 addressed)

Closes #77, closes #107

NB: merge #108 and #110 first

[johnlees]

97d215f closes #111

[johnlees]

TODO: change --fit-model to take options gmm, dbscan, refine, threshold or use

[johnlees]

Also, remove the --full-db option from assign mode

[nickjcroucher]

All looks great - made a couple of suggestions on how the command line instructions might be simplified in the context of these changes, would be interested to know what you think.

[johnlees]

Suggestions addressed in #125