fit_phi_mech: Fit intervention-specific exponential tilt nuisance parameter
In nhejazi/medshift: Causal mediation analysis for stochastic interventions

Fit intervention-specific exponential tilt nuisance parameter

1	fit_phi_mech(train_data, valid_data, learners, m_output, w_names)

`train_data`	A `data.table` containing the observed data, with columns in the order specified by the NPSEM (Y, Z, A, W), with column names set appropriately based on the input data. Such a structure is merely a convenience utility to passing data around to the various core estimation routines and is automatically generated by `medshift`.
`valid_data`	A holdout data set, with columns exactly matching those appearing in the preceding argument `train_data`, to be used for estimation via cross-fitting. Not optional for this nuisance parameter.
`learners`	A `Stack` (or other learner class that inherits from `Lrnr_base`), containing a single or set of instantiated learners from sl3, to be used in a regression of a pseudo-outcome on the baseline covariates, i.e., phi(W) = E[m(A = 1, Z, W) - m(A = 0, Z, W) \| W).
`m_output`	Object containing results from fitting the outcome regression, as produced by `fit_m_mech`.
`w_names`	A `character` vector of the names of the columns that correspond to baseline covariates (W). The input for this argument is automatically generated by `medshift`.