Allow editing of `Node` production/consumption Amounts... to update the LCA Result

[michaelweinold]

Currently, the dashboard returns a list of nodes in the supply chain tree like so:

UID	Activity Description	Supply Amount	Branch
1	Cardboard	100	1
2	Paper	45	[1,2]
3	Wood Pulp	60	[1,2,3]
4	Electricity	110	[1,2,3,4]

The user should be able to say:

"I know we actually use only 30kg of Paper to make 100kg of Cardboard!"

and then edit this Supply Amount in the second row.

Of course, this would mean that correspondingly less Supply Amount of all activities feeding into Paper would be required. Of course, all this would also affect the overall LCA score.

[michaelweinold]

...maybe there is a way propagate edits from the table back to the dict of dicts, which is returned by NewNodeEachVisitGraphTraversal.calculate(lca, cutoff=0.01). Would this make it easier to re-compute upstream Supply Amounts and the overall score?

{'nodes': {-1: Node(unique_id=-1, activity_datapackage_id=-1, activity_index=-1, reference_product_datapackage_id=-1, reference_product_index=-1, reference_product_production_amount=1.0, depth=0, supply_amount=1.0, cumulative_score=0.3768834121437975, direct_emissions_score=0.0, direct_emissions_score_outside_specific_flows=0.0, remaining_cumulative_score_outside_specific_flows=0.0, terminal=False),
  0: Node(unique_id=0, activity_datapackage_id=235, activity_index=234, reference_product_datapackage_id=542, reference_product_index=153, reference_product_production_amount=0.9988390803337097, depth=1, supply_amount=1.0011622689671917, cumulative_score=0.3768834121437976, direct_emissions_score=0.015464815391143152, direct_emissions_score_outside_specific_flows=1.0495129317330404e-05, remaining_cumulative_score_outside_specific_flows=0.3614290918819718, terminal=False),
  1: Node(unique_id=1, activity_datapackage_id=79, activity_index=78, reference_product_datapackage_id=431, reference_product_index=42, reference_product_production_amount=0.9359190464019775, depth=2, supply_amount=0.013377550027867835, cumulative_score=0.004254062537767607, direct_emissions_score=0.00012451661922425745, direct_emissions_score_outside_specific_flows=3.694040387367938e-06, remaining_cumulative_score_outside_specific_flows=0.004133239958930717, terminal=True),
  2: Node(unique_id=2, activity_datapackage_id=311, activity_index=310, reference_product_datapackage_id=435, reference_product_index=46, reference_product_production_amount=0.9928678274154663, depth=2, supply_amount=0.09230018343308832, cumulative_score=0.041779131103952524, direct_emissions_score=0.001482884746645427, direct_emissions_score_outside_specific_flows=4.342210964385475e-05, remaining_cumulative_score_outside_specific_flows=0.04033966846695095, terminal=False),
(...)
 'edges': [Edge(consumer_index=-1, consumer_unique_id=-1, producer_index=234, producer_unique_id=0, product_index=153, amount=1),
  Edge(consumer_index=234, consumer_unique_id=0, producer_index=78, producer_unique_id=1, product_index=42, amount=0.012520303865276812),
  Edge(consumer_index=234, consumer_unique_id=0, producer_index=310, producer_unique_id=2,
(...)
 'flows': [Flow(flow_datapackage_id=1726, flow_index=947, activity_unique_id=19, activity_id=374, activity_index=373, amount=0.016917697121138834, score=0.016917697121138834),
  Flow(flow_datapackage_id=1726, flow_index=947, activity_unique_id=3, activity_id=301, activity_index=300, amount=0.013172648927940041, score=0.013172648927940041),
  Flow(flow_datapackage_id=1726, flow_index=947, activity_unique_id=4, activity_id=53, activity_index=52, amount=0.007941925100944729, score=0.007941925100944729),
(...)
}

[cmutel]

Any time you change the matrix you need to start graph traversal again. It will break the matrix factorization cache, and change all LCA results.

[cmutel]

I would do this by keeping a separate datapackage in memory, following https://github.com/brightway-lca/from-the-ground-up/blob/main/2%20-%20Building%20and%20using%20matrices%20in%20bw2calc.ipynb, as then you don't need to touch the SQLite database.