Package index

Generics

The S3 generics the r-causal ecosystem shares. Packages register methods on these rather than defining their own, so attaching several at once produces no masking.

ipw()

Inverse probability weighted estimation

as_marginal() as_conditional()

The presentation mode of an inverse probability weighted result

ess()

Effective sample size

is_causal_wt() estimand() `estimand<-`()

Causal weight accessors

Shared classes

Constructors and helpers for the classes those generics return. Method authors build on these instead of writing a result or weight class of their own.

new_ipw() print(<ipw>) as.data.frame(<ipw>)

Construct an inverse probability weighted result

new_ipw_model() vcov(<ipw_model>)

Attach a corrected covariance to a component model of an IPW fit

coef(<ipw>) vcov(<ipw>) confint(<ipw>) nobs(<ipw>) df.residual(<ipw>) weights(<ipw>) model.frame(<ipw>) estimand(<ipw>)

Model accessors for an inverse probability weighted result

pool_ipw()

Pool inverse probability weighted results over multiply imputed data

print(<ipw_pooled>) coef(<ipw_pooled>) vcov(<ipw_pooled>) confint(<ipw_pooled>) nobs(<ipw_pooled>) as.data.frame(<ipw_pooled>) estimand(<ipw_pooled>)

Methods for a pooled inverse probability weighted result

new_causal_wts()

Construct a causal weight vector

causal_wts_ptype2()

Common type of two causal weight vectors of the same class

Exposure declarations

Classes for the exposure an analysis is estimated from rather than for what it returns. A package handed such a column reads the declaration off it instead of inferring the design from the values it holds.

joint_exposure()

Declare an exposure that crosses two treatments

is_joint_exposure() joint_components() joint_reference()

Read the declaration a joint exposure carries

Exposure types

The reading of whether an exposure is binary, categorical, or continuous, and the resolution of the type a caller declared against the types a function supports. Packages take the reading from here so that the same column is treated the same way wherever it is analyzed.

detect_exposure_type()

Detect the type of an exposure

match_exposure_type()

Resolve the exposure type a function will work on

check_forced_type()

Refuse an exposure type the data cannot carry

observed_values() has_two_levels() is_categorical()

Read the values an exposure takes