Incorporate beta test feedback and prepare for counterfeit pills analysis

[sandypreiss]

This PR incorporates changes suggested by Mark Edlund, Paolo Manelli, and Dan Ciccarone during the beta testing process.

Changes include:

• Adding a dose translation table, so the user specifies dose in terms of a specific opioid, which is then converted to MMEs
• Add the concept of counterfeit pills and adjust variability mechanisms accordingly
  • Add counterfeit pills slider
  • Dose variability and fentanyl adulteration now affect counterfeit doses only
  • Add behavioral variability to reflect intentional variation of dose as well as variation caused by uncertain concentration of counterfeit pills
• Improve overdose risk calculation
  • Calculate baseline risk from Dasgupta et al
  • Assume baseline risk is for tolerant users, add risk multiplier when user takes dose above their tolerance
• Calibrate decay function to morphine plasma concentration half life
• Enforce 2 doses per day and calibrate tolerance function to produce desired dose increase behavior given 2-dose paradigm