adding type specs to jitted functions?

[amoodie]

Above the _get_water_weight_array() function there's commented out code specifying the expected input types in the @njit decorator.

As far as I understand it, adding the types does not speed up the actual computations, but it does speed up the actual just-in-time compilation. So instead of the first timestep taking 30 seconds, and then the next ones taking 1 second, the first could be reduced to ~5 seconds if the types are specified.

This is nice, but it is definitely a confusing and intimidating block of code for someone new to python/numba who may want to implement their own methods/functions or modify our own. I think it's something to consider, and with the appropriate documentation, may be worthwhile.