consider having the Jacobian type depend on T

[zingale]

For the subchandra problem, we run much faster with the analytic Jacobian, but for really high temperatures at the point of the C det ignition, we sometimes need to use the numerical Jacobian. It would be interesting to be able to switch to the numerical Jacobian based on T when entering the burn.

[zingale]

removing this now, since we have the Jacobian swap on retry option