pair_trans: Generate covariates to fit the logistic regression, which is...
In chenchi0526/SIsMiss:

pair_trans is used to generate transformed covariates in pairwise form.

1	pair_trans(y, z, u, eta_est, z.sigma_est)

`y`	Vector. The response variable subject to missing data.
`z`	Vector. The shadow variable (fully observed).
`u`	Matrix. The covariate matrix excluding shadow variable z (fully observed).
`eta_est`	The estimates for coefficients in model Z givenU.
`z.sigma_est`	The estimates for σ in model Z givenU.

A list:

data_pair: The matrix of pairwised y, z, u.
newcov_pair: The matrix of pairwised transformed variables to be fit in the objective function.
obs_id: The indice of missingness for subjects in data.
w_hat: The estimates of W matrix in the offset term.