Function adjustBackdoor

adjustBackdoor(
    graph: CausalGraph,
    treatment: string,
    outcome: string,
    data: number[][],
    nodeIndex: Map<string, number>,
): { adjustors: string[]; ate: number; se: number }
Compute ATE using backdoor adjustment.

ATE = E[ Y|do(X=x₁) ] - E[ Y|do(X=x₀) ] For linear models: β_treatment from OLS(Y ~ X + Z)
Parameters
- graph: CausalGraph
- treatment: string
- outcome: string
- data: number[][]
- nodeIndex: Map<string, number>
Returns { adjustors: string[]; ate: number; se: number }
- Defined in pipeline/src/infer/effect-estimation.ts:56