Add top parameter to loec fitting

[brown-jason]

Top is used downstream in a few different cases so i think it's important to specify a top.
Per 12/10 standup from Sarah
There may be cases that the LOEC isn't actual the maximum change in response from the baseline. That is if you have 5 concentration groups and say your LOEC is at conc group 3 and 4 and 5 have larger median responses, then top assigned at the loec median response is not particularly true to how that estimated value is typically defined across ToxCast.

[brown-jason]

Top should be defined as the maximum median response in the direction of the fit LOEC.