Skip to contents

Creates IPTW balancing weights

Source: R/createWeights.R
createWeights.Rd

Creates IPTW balancing weights at each user-specified exposure time point using balancing formulas that relate exposure at each time point to all relevant confounders.

Usage

createWeights(
  data,
  formulas,
  method = "glm",
  verbose = FALSE,
  save.out = FALSE,
  ...
)

# S3 method for class 'devMSM_weights'
print(x, i = 1, ...)

# S3 method for class 'devMSM_weights'
plot(x, i = 1, save.out = FALSE, ...)

# S3 method for class 'devMSM_weights'
summary(object, i = 1, ...)

Arguments

data

data in wide format as: a data frame, list of imputed data frames, or mids object from the mice package

formulas

list of balancing formulas at each time point output from createFormulas()

method

character string of weightitMSM() balancing method abbreviation (default is generalized linear models propensity score weighting "glm")

verbose

(optional) TRUE or FALSE indicator for printing output to console. default is FALSE.

save.out

(optional) Either logical or a character string. If TRUE, it will output the result to a default file name within home_dir set in initMSM(). You can load the data with x <- readRDS(file). To use a non-default file name, specify a character string with the file name. It will save relative to home_dir. There might be naming conflicts where two objects get saved to the same file. In these cases, users should specify a custom name. default is FALSE.

...

arguments passed to WeightIt::weightitMSM() or summary.weightitMSM()

x, object

devMSM_weights object from createWeights()

i

For multiply imputed datasets, i selects which imputation to print results for. Default is i = 1. With i = TRUE, all imputed datasets will be looped over. With i = NULL, will average over all imputed datasets and summarize that. Ignored for non-imputed data.

Value

a list containing WeightIt::weightitMSM() output. It is the length of the number of datasets (1 for a data.frame or the number of imputed datasets).

See also

WeightIt::weightitMSM()

Examples

library(devMSMs)
data <- data.frame(
  ID = 1:50,
  A.1 = rnorm(n = 50),
  A.2 = rnorm(n = 50),
  A.3 = rnorm(n = 50),
  B.1 = rnorm(n = 50),
  B.2 = rnorm(n = 50),
  B.3 = rnorm(n = 50),
  C = rnorm(n = 50),
  D.3 = rnorm(n = 50)
)
obj <- initMSM(
  data,
  exposure = c("A.1", "A.2", "A.3"),
  ti_conf = c("C"),
  tv_conf = c("B.1", "B.2", "B.3", "D.3")
)
f <- createFormulas(obj, type = "short")

w <- createWeights(data = data, formulas = f)
print(w)
#> 
#> For the `glm` weighting method, the median weight value is 1.04 (SD = 0.41; range = 0.3-3).
plot(w)


# Methods from `WeightIt::weightitMSM`
w <- createWeights(data = data, formulas = f,
                   method = "glm")

w <- createWeights(data = data, formulas = f,
                   method = "cbps")
w <- createWeights(data = data, formulas = f,
                   method = "gbm")
w <- createWeights(data = data, formulas = f,
                   method = "bart")
w <- createWeights(data = data, formulas = f,
                   method = "super")
#> Loading required package: nnls
#> Warning: All algorithms have zero weight
#> Warning: All metalearner coefficients are zero, predictions will all be equal to 0