As districts tune the optimization, we want to be able to customize the strategy parameters to try to find an optimal result.
The idea would be to have a "Custom Optimization" in a somewhat obfuscated (e.g. advanced collapse area) to allow specifying the string specification of a strategy (e.g. NYCMODA?iterations=1200&fresh_starts=100). At this point we want this to be more a part of the "support" workflow where someone reaches out to the team and we provide them the parameters we see as getting a best result for their goals. Eventually it would be good to allow this (with "limits" to prevent mammoth AWS bills) for people to experiment with? TBD
This can then be parsed by server.py and turned into the strategies list that is sent off to lambda (not recommended for immediate processing!)
As districts tune the optimization, we want to be able to customize the strategy parameters to try to find an optimal result.
The idea would be to have a "Custom Optimization" in a somewhat obfuscated (e.g. advanced collapse area) to allow specifying the string specification of a strategy (e.g. NYCMODA?iterations=1200&fresh_starts=100). At this point we want this to be more a part of the "support" workflow where someone reaches out to the team and we provide them the parameters we see as getting a best result for their goals. Eventually it would be good to allow this (with "limits" to prevent mammoth AWS bills) for people to experiment with? TBD
This can then be parsed by server.py and turned into the strategies list that is sent off to lambda (not recommended for immediate processing!)