strain introduction lists mirror strain enum.

[arik-shurygin]

the current system is best shown through an example. lets say we have the following config parameters:

{
"STRAIN_IDX": ["STRAINX", "STRAINY", "STRAINZ"],
"STRAIN_R0s": [1.2, 1.8, 3.0],
"INTRODUCTION_TIMES": [60, 100],
"INTRODUCTION_PCTS": [0.01, 0.02],
"INTRODUCTION_SCALES": [10, 10],
}

It is fairly clear from the example that there are two strains being introduced here, but which ones, first two or last two? Those who use DynODE already know that it is STRAINY and STRAINZ but this is not immediately obvious. Lets revamp this system so that it is more clear to someone learning DynODE how to introduce strains.

Proposed schema:

{
"STRAIN_IDX": ["STRAINX", "STRAINY", "STRAINZ"],
"STRAIN_R0s": [1.2, 1.8, 3.0],
"INTRODUCTION_FLAG": [false, true, true]
"INTRODUCTION_TIMES": [-1, 60, 100],
"INTRODUCTION_PCTS": [0.0, 0.01, 0.02],
"INTRODUCTION_SCALES": [0, 10, 10],
}

this way it is explicitly shown to the user which strain is being introduced and which are not. It also grants users the ability to introduce STRAINX and STRAINZ but NOT STRAINY which is currently not possible.

This will likely require the following changes:

• in order to ignore filler values placed by the user in non-introduced strains, some changes to the validate functions of the existing introduction parameters will be required within config.py
• the abstract_parameters.external_i() function will also need to read this INTRODUCTION_FLAG, but this should be fairly easy to implement.

[kokbent]

looks good to me... in theory negative intro time could have meant the strain is not introduced... then abstract_parameters.external_i() wouldn't need to read INTRODUCTION_FLAG but I guess being more explicit is good and you could actually have introduction time of -1 in our model and that makes things a bit complicated.