Build a validated dp object from raw user inputs

Build a validated dp object from raw user inputs

Usage

make_did_dp(
  data = NULL,
  y,
  d,
  post,
  covariates = NULL,
  dim_covariates = NULL,
  continuous = NULL,
  unordered = NULL,
  ordered = NULL,
  na_action = c("fail", "omit"),
  validate = TRUE
)

Arguments

data: Optional data frame or matrix containing the analysis variables.
y: Outcome vector, or a single column name/index when data is supplied.
d: Binary treatment indicator, or a single column name/index when data is supplied.
post: Binary post-period indicator, or a single column name/index when data is supplied.
covariates: Optional preassembled covariate matrix. If omitted, the covariates are built from data using the block selectors below.
dim_covariates: Optional length-3 dimension vector. This must be supplied when covariates is already assembled and the block selectors are not used.
continuous: Character or integer vector identifying continuous covariates in data or in covariates.
unordered: Character or integer vector identifying unordered discrete covariates in data or in covariates.
ordered: Character or integer vector identifying ordered discrete covariates in data or in covariates.
na_action: How to handle missing values in the variables used to build dp. "fail" stops, and "omit" drops incomplete rows.
validate: Logical; if TRUE, validate the output with validate_did_dp().

Value

A list in the internal dp format.