possibility to extract sugbraphs from complex models

[dberardo-com]

i have a quite complex excel model, that contains a lot of subgraphs with different roots. in order to reduce computational complexity i would like to isolate only a selection of subgraphs, ignoring all others.

is it possible to pick one subgraph starting from a root node and discard all the others ?

[dberardo-com]

ok i have found out this as a way to get subgraphs:

model=xl_model
model.dsp.dispatch(inputs=final_inps, no_call=True)
sub_disp=model.dsp.get_sub_dsp_from_workflow(sources=outputs.values(), reverse=True)

this seems to work but my questions are now:

• if i call sub_disp.dispatch() is that in any way more efficient than model.compute() ? or is it going to be the same ?
• i have seen that loading the excel model takes up 2GB of RAM during the call to .finish() . but i have seen that if i call only loads().complete(), without .assemble() then i can still get the full subgraph displayed using .plot() function. so this makes me think that it should be possible to .assemble() only the subgraphs without having to first assemble() the whole model. this would make me save some memory i guess. is this possible ?

[dberardo-com]

in particular i have figured out that the memory usage is quite big when loading the VLOOKUP excel function. how so ? can that be improved ?