refine: add argument for divergence unit

[rneher]

this adds and argument --divergence-unit to refine with choices mutations and mutations-per-site (default). Thought was we might have a case with amino-acids or similar so keeping it general.

[trvrb]

I really like this implementation. Testing worked well. I'm going to merge and push augur release.

[hackerchenzhuo]

good idea. by the way , what does it mean when it is default during divergence computing?

[tsibley]

@hackerchenzhuo

what does it mean when it is default during divergence computing?

I assume you're asking about this bit of code:

augur/augur/refine.py

Lines 223 to 224 in 91f82f5

	if args.divergence_units=='mutations-per-site': #default
	pass

Internally divergence is always computed as mutations-per-site. If that's the desired output unit, then nothing needs to be done to convert the units before outputting. This is also the default output unit, meaning that if --divergence-unit mutations is not specified to augur refine, then it'll use mutations-per-site.

[hackerchenzhuo]

@hackerchenzhuo

what does it mean when it is default during divergence computing?

I assume you're asking about this bit of code:

augur/augur/refine.py

Lines 223 to 224 in 91f82f5

	if args.divergence_units=='mutations-per-site': #default
	pass

Internally divergence is always computed as mutations-per-site. If that's the desired output unit, then nothing needs to be done to convert the units before outputting. This is also the default output unit, meaning that if --divergence-unit mutations is not specified to augur refine, then it'll use mutations-per-site.

hi @tsibley
thank you for your reply
I have another question: can we comput the divergence between any two nodes? which means comput from any two different gene sequence rather than from root to other nodes like what discribe in algorithm?
Or maybe I have the wrong understand.