make it easy to interface with outside solvers
Closed this issue · 2 comments
rmodrak commented
Gian Matharu has interfaced SeisFlows with a 2D finite difference solver. Starting to look through Gian's code for ideas on how to make it easy to work with external solvers.
rmodrak commented
So far, made some simple changes to the inversion
workflow. I did, however, leave in the initialize
method, which is useful, for example, if you want to implement source encoding or source decimation. See also thrifty_inversion
workflow, which uses on initialize
and, at the expense of some complexity, provides quite significant computational savings.
rmodrak commented
Next, will explore further if it is possible to decouple the data reader from the solver in the way Gian was suggesting.