Package 'medshift'

An object of class medshift, as produced by invoking medshift, for which a confidence interval is to be computed.

A numeric vector indicating indices of object$est for which to return confidence intervals.

A numeric indicating the confidence interval level.

Other arguments. Not currently used.

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 original 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.

A numeric value indicating the degree of shift in the intervention to be used in defining the causal quantity of interest. In the case of binary interventions, this takes the form of an incremental propensity score shift, acting as a multiplier of the odds with which a given observational unit receives the intervention (EH Kennedy, 2018, JASA; https://doi.org/10.1080/01621459.2017.1422737).

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 fitting the propensity score, i.e., g = P(A | W).

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 fitting a propensity score that conditions on the mediators, i.e., e = P(A | Z, W).

A character vector of the names of the columns that correspond to baseline covariates (W). The input for this argument is automatically generated by a call to the wrapper function medshift.

A character vector of the names of the columns that correspond to mediators (Z). The input for this argument is automatically generated by medshift.

Other arguments currently ignored.

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 original 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.

A numeric value indicating the degree of shift in the intervention to be used in defining the causal quantity of interest. In the case of binary interventions, this takes the form of an incremental propensity score shift, acting as a multiplier of the probability with which a given observational unit receives the intervention (EH Kennedy, 2018, JASA; https://doi.org/10.1080/01621459.2017.1422737).

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 fitting the propensity score, i.e., g = P(A | W).

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 fitting a propensity score that conditions on the mediators, i.e., e = P(A | Z, W).

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 fitting the outcome regression, i.e., m(A, Z, W).

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).

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.

A character vector of the names of the columns that correspond to mediators (Z). The input for this argument is automatically generated by medshift.

A numeric specifying the number of folds to be created for cross-validation. Use of cross-validation / cross-fitting allows for entropy conditions on the AIPW estimator to be relaxed. Note: for compatibility with make_folds, this value must be greater than or equal to 2; the default is to create 10 folds.

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 original 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.

A numeric value indicating the degree of shift in the intervention to be used in defining the causal quantity of interest. In the case of binary interventions, this takes the form of an incremental propensity score shift, acting as a multiplier of the odds with which a given observational unit receives the intervention (EH Kennedy, 2018, JASA; https://doi.org/10.1080/01621459.2017.1422737).

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 fitting the propensity score, i.e., g = P(A | W).

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 fitting the outcome regression, i.e., m(A, Z, W).

A character vector of the names of the columns that correspond to baseline covariates (W). The input for this argument is automatically generated by a call to the wrapper function medshift.

A character vector of the names of the columns that correspond to mediators (Z). The input for this argument is automatically generated by a call to the wrapper function medshift.

Other arguments currently ignored.

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 original 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.

A holdout data set, with columns exactly matching those appearing in the preceding argument data, to be used for estimation via cross-fitting. Optional, defaulting to NULL.

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 fitting a propensity score that conditions on the mediators, i.e., e = P(A | Z, W).

A character vector of the names of the columns that correspond to mediators (Z). The input for this argument is automatically generated by medshift.

A character vector of the names of the columns that correspond to baseline covariates (W). The input for this argument is automatically generated by a call to the wrapper function medshift.

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 original 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.

A holdout data set, with columns exactly matching those appearing in the preceding argument data, to be used for estimation via cross-fitting. Optional, defaulting to NULL.

A numeric value indicating the degree of shift in the intervention to be used in defining the causal quantity of interest. In the case of binary interventions, this takes the form of an incremental propensity score shift, acting as a multiplier of the odds with which a given observational unit receives the intervention (EH Kennedy, 2018, JASA; https://doi.org/10.1080/01621459.2017.1422737).

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 fitting the propensity score, i.e., g = P(A | W).

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.

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 original 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.

A holdout data set, with columns exactly matching those appearing in the preceding argument data, to be used for estimation via cross-fitting. Optional, defaulting to NULL.

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 fitting the outcome regression, i.e., m(A, Z, W).

A character vector of the names of the columns that correspond to mediators (Z). The input for this argument is automatically generated by medshift.

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.

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.

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.

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).

Object containing results from fitting the outcome regression, as produced by fit_m_mech.

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.

A matrix, data.frame, or similar corresponding to a set of baseline covariates.

A numeric vector corresponding to a treatment variable. The parameter of interest is defined as a location shift of this quantity.

A numeric vector, matrix, data.frame, or similar corresponding to a set of mediators (on the causal pathway between the intervention A and the outcome Y).

A numeric vector corresponding to an outcome variable.

A numeric vector of observation-level IDs, allowing for observational units to be related through a hierarchical structure. The default is to assume all units are IID. When repeated IDs are included, both the cross-validation procedures used for estimation and inferential procedures respect these IDs.

A numeric value indicating the degree of shift in the intervention to be used in defining the causal quantity of interest. In the case of binary interventions, this takes the form of an incremental propensity score shift, acting as a multiplier of the odds with which a unit receives the intervention (EH Kennedy, 2018, JASA; https://doi.org/10.1080/01621459.2017.1422737).

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 fitting the propensity score, i.e., g = P(A | W).

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 fitting a propensity score that conditions on the mediators, i.e., e = P(A | Z, W).

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 fitting the outcome regression, i.e., m(A, Z, W).

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).

The desired estimator of the natural direct effect to be computed. Currently, choices are limited to a substitution estimator, a re-weighted estimator, a one-step estimator, and a targeted minimum loss estimator.

A list of extra arguments to be passed (via ...) to the function call for the specified estimator. The default is so chosen as to allow the number of folds used in computing the one-step estimator to be easily tweaked. Refer to the documentation for functions est_onestep, est_ipw, and est_substitution for details on what other arguments may be specified through this mechanism. For the option "tmle", there is heavy reliance on the architecture provided by tmle3.

A matrix, data.frame, or similar corresponding to a set of baseline covariates.

A numeric vector corresponding to a treatment variable. The parameter of interest is defined as a location shift of this quantity.

A numeric vector, matrix, data.frame, or similar corresponding to a set of mediators (on the causal pathway between the intervention A and the outcome Y).

A numeric vector corresponding to an outcome variable.

A numeric vector of observation-level IDs, allowing for observational units to be related through a hierarchical structure. The default is to assume all units are IID. When repeated IDs are included, both the cross-validation procedures used for estimation and inferential procedures respect these IDs.

A numeric value indicating the degree of shift in the intervention to be used in defining the causal quantity of interest. In the case of binary interventions, this takes the form of an incremental propensity score shift, acting as a multiplier of the odds with which a unit receives the intervention (EH Kennedy, 2018, JASA; https://doi.org/10.1080/01621459.2017.1422737).

The desired estimator of the natural direct effect to be computed. Currently, choices are limited to a substitution estimator, a re-weighted estimator, a one-step estimator, and a targeted minimum loss estimator.

A numeric indicating the desired coverage level of the confidence interval to be computed.

Additional arguments passed to medshift. Consult the documentation of that function for details.

An object of class medshift.

Other options (not currently used).

An object of class medshift, as produced by invoking the function medshift, for which a confidence interval is to be constructed.

Other arguments. Not currently used.

A numeric indicating the level of the confidence interval to be computed.

A matrix, data.frame, or similar corresponding to a set of baseline covariates.

A numeric vector corresponding to a treatment variable. The parameter of interest is defined as a location shift of this quantity.

A numeric vector, matrix, data.frame, or similar corresponding to a set of mediators (on the causal pathway between the intervention A and the outcome Y).

A numeric vector corresponding to an outcome variable.

A numeric vector of observation-level IDs, allowing for observational units to be related through a hierarchical structure. The default is to assume all units are IID. When repeated IDs are included, both the cross-validation procedures used for estimation and inferential procedures respect these IDs.

A numeric of values giving the varous degrees of shift in the intervention to be used in defining the causal quantity of interest. In the case of binary interventions, this takes the form of an incremental propensity score shift, acting as a multiplier of the odds with which a given observational unit receives the intervention (EH Kennedy, 2018, JASA; https://doi.org/10.1080/01621459.2017.1422737).

A character identifying the type of multipliers to be used in the multiplier bootstrap. Choices are "rademacher" or "gaussian", with the default being the former.

A numeric indicating the (1 - alpha) level of the simultaneous confidence band to be computed around the estimates of the direct effect. The error level of the test reported in the p-value returned is simply alpha, i.e., one less this quantity.

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 fitting the propensity score, i.e., g = P(A | W).

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 fitting a propensity score that conditions on the mediators, i.e., e = P(A | Z, W).

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 fitting the outcome regression, i.e., m(A, Z, W).

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).

A numeric specifying the number of folds to be created for cross-validation. Use of cross-validation / cross-fitting allows for entropy conditions on the AIPW estimator to be relaxed. Note: for compatibility with make_folds, this value must be greater than or equal to 2; the default is to create 10 folds.

A numeric scalar giving the number of repetitions of the multipliers to be used in computing the multiplier bootstrap.

A character defining the type of shift to be applied to the exposure – an incremental propensity score intervention, by default.

A numeric, specifying the magnitude of the shift.

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 fitting a cleverly parameterized propensity score that conditions on the mediators, i.e., e = P(A | Z, W).

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).

A numeric setting the maximum iterations allowed in the targeting step based on universal least favorable submodels.

A numeric giving the step size (delta_epsilon in tmle3) to be used in the targeting step based on universal least favorable submodels.

Additional arguments (currently unused).