Remove use of CausalModel from notebooks and test files

[rahulbshrestha]

This PR addresses this issue. Since CausalModel is deprecated, it should be removed from tutorials and test cases.

[rahulbshrestha]

@bloebp I did it for one file, could you please let me know if this is fine?

The only thing is when I tried to compare the variables for identify_effect for before and after, it seems to be different? e.g if I do

from dowhy.causal_identifier import identify_effect
identification_1 = identify_effect(nx_graph, action_nodes=causes, outcome_nodes=outcomes, observed_nodes=list(graph.get_all_nodes(include_unobserved=False)))

model = CausalModel(df, causes, outcomes, common_causes=common_causes),
nx_graph = model._graph._graph
identification_2 = model.identify_effect(proceed_when_unidentifiable=True)

identification_1 == identification_2

The output is False, any idea why? I can dig into this a bit more too.

[amit-sharma]

thanks for starting this @rahulbshrestha
In what way does the two identification objects differ? One difference I see is that you may need to use nx_graph as input to CausalModel too (causalmodel accepts a nx graph as graph)

[rahulbshrestha]

@amit-sharma I investigated this a bit more and found the difference between both:

identification_1.identifier returns None

identification_2.identifier returns <dowhy.causal_identifier.auto_identifier.AutoIdentifier at 0x2876e2690>

Is this an intended design choice or a bug?

[bloebp]

Hey, I think @amit-sharma is better in answering this. Generally, I would expect that we obtain 1:1 the same. If not, there might be an issue we should investigate.

[amit-sharma]

@rahulbshrestha yeah you are right! I checked the two implementations of identify_effect. The older 'CausalModelimplementation returns a reference to theCausalIdentifier` class that computed the identification. But this does not apply for the new API since it is function-based.

So we'll need to implement a custom __eq__ method for the IdentifiedEstimand class that only compares all the remaining attributes. Would you be up for adding that?