sa {CIMTx}

Flexible Monte Carlo sensitivity analysis for unmeasured confounding

Description

The function sa implements the flexible sensitivity analysis approach for unmeasured confounding with multiple treatments and a binary outcome.

Usage

sa(
  x,
  y,
  w,
  formula = NULL,
  prior_c_function,
  m1,
  m2 = NULL,
  n_cores = 1,
  estimand,
  reference_trt,
  ...
)

Arguments

x

A dataframe, including all the covariates but not treatments.

y

A numeric vector (0, 1) representing a binary outcome.

w

A numeric vector representing the treatment groups.

formula

A formula object for the analysis. The default is to use all terms specified in x.

prior_c_function

1) A vector of characters indicating the prior distributions for the confounding functions. Each character contains the random number generation code from the standard probability distributions in the stats package. 2) A vector of characters including the grid specifications for the confounding functions. It should be used when users want to formulate the confounding functions as scalar values. 3) A matrix indicating the point mass prior for the confounding functions

m1

A numeric value indicating the number of draws of the GPS from the posterior predictive distribution

m2

A numeric value indicating the number of draws from the prior distributions of the confounding functions

n_cores

A numeric value indicating number of cores to use for parallel computing.

estimand

A character string representing the type of causal estimand. Only "ATT" or "ATE" is allowed. When the estimand = "ATT", users also need to specify the reference treatment group by setting the reference_trt argument.

reference_trt

A numeric value indicating reference treatment group for ATT effect.

...

Other parameters that can be passed to BART functions

Value

A list of causal estimands including risk difference (RD) between different treatment groups.

References

Hadley Wickham (2019). stringr: Simple, Consistent Wrappers for Common String Operations. R package version 1.4.0. URL:https://CRAN.R-project.org/package=stringr

Hadley Wickham (2021). tidyr: Tidy Messy Data. R package version 1.1.4. URL:https://CRAN.R-project.org/package=tidyr

Sparapani R, Spanbauer C, McCulloch R Nonparametric Machine Learning and Efficient Computation with Bayesian Additive Regression Trees: The BART R Package. Journal of Statistical Software, 97(1), 1-66.

Microsoft Corporation and Steve Weston (2020). doParallel: Foreach Parallel Adaptor for the 'parallel' Package. R package version 1.0.16. URL:https://CRAN.R-project.org/package=doParallel

Microsoft and Steve Weston (2020). foreach: Provides Foreach Looping Construct.. R package version 1.5.1 URL:https://CRAN.R-project.org/package=foreach

Examples

lp_w_all <-
  c(
    ".4*x1 + .1*x2  - 1.1*x4 + 1.1*x5", # w = 1
    ".2 * x1 + .2 * x2  - 1.2 * x4 - 1.3 * x5"
  ) # w = 2
nlp_w_all <-
  c(
    "-.5*x1*x4  - .1*x2*x5", # w = 1
    "-.3*x1*x4 + .2*x2*x5"
  ) # w = 2
lp_y_all <- rep(".2*x1 + .3*x2 - .1*x3 - 1.1*x4 - 1.2*x5", 3)
nlp_y_all <- rep(".7*x1*x1  - .1*x2*x3", 3)
X_all <- c(
  "rnorm(0, 0.5)", # x1
  "rbeta(2, .4)", # x2
  "runif(0, 0.5)", # x3
  "rweibull(1,2)", # x4
  "rbinom(1, .4)" # x5
)
set.seed(1111)
data <- data_sim(
  sample_size = 100,
  n_trt = 3,
  x = X_all,
  lp_y = lp_y_all,
  nlp_y = nlp_y_all,
  align = FALSE,
  lp_w = lp_w_all,
  nlp_w = nlp_w_all,
  tau = c(0.5, -0.5, 0.5),
  delta = c(0.5, 0.5),
  psi = 2
)
c_grid <- c(
  "runif(-0.6, 0)", # c(1,2)
  "runif(0, 0.6)", # c(2,1)
  "runif(-0.6, 0)", # c(2,3)
  "seq(-0.6, 0, by = 0.3)", # c(1,3)
  "seq(0, 0.6, by = 0.3)", # c(3,1)
  "runif(0, 0.6)" # c(3,2)
)
sensitivity_analysis_parallel_result <-
  sa(
    m1 = 1,
    x = data$covariates,
    y = data$y,
    w = data$w,
    prior_c_function = c_grid,
    n_cores = 1,
    estimand = "ATE",
  )