method_ps | R Documentation |
Function to specify the propensity score (PS) model for the inverse probability weighting estimator.
This function provides basic functions logistic regression with a given link function (currently
we support logit
, probit
and cloglog
) with additional information about the analytic variance estimator of the mean.
This is a function returns a list
of functions that refer to specific estimation methods and variance estimators
when whether the IPW alone or the DR estimator is applied. The export of this function is mainly because
the functions are used in the variable selection algorithms.
Functions starting with make_log_like
, make_gradient
and make_hessian
refer to the maximum likelihood estimation
as described in the Chen et al. (2020) paper. These functions take into account different link functions defined through
the link
argument.
Functions make_link
, make_link_inv
, make_link_der
, make_link_inv_der
, make_link_inv_rev
, and make_link_inv_rev_der
refer to specific link functions and are used in the estimation process.
Functions variance_covariance1
and variance_covariance2
refer to the variance estimator of the IPW estimator as
defined by Chen et al. (2020).
Functions b_vec_ipw
, b_vec_dr
and t_vec
are specific functions defined in the Chen et al. (2020) that are used
in the variance estimator of the IPW or the DR.
Finally, var_nonprob
is the non-probability component of the DR estimator as defined by Chen et al. (2020).
method_ps(link = c("logit", "probit", "cloglog"), ...)
link |
link for the PS model |
... |
Additional, optional arguments. |
A list
of functions and elements for a specific link function with the following entries:
log-likelihood function for a specific link function
gradient of the loglik
hessian of the loglik
link function
inverse link function
first derivative of the link function
first derivative of the the inverse link function
defines 1/inv_link
first derivative of 1/inv_link
for the IPW estimator: variance component for the non-probability sample
for the IPW estimator: variance component for the probability sample
for the IPW estimator: the b
function as defined in the Chen et al. (2020, sec. 3.2, eq. (9)-(10); sec 4.1)
for the DR estimator: the b
function as defined in the Chen et al. (2020, sec. 3.3., eq. (14); sec 4.1)
for the DR estimator: the b
function as defined in the Chen et al. (2020, sec. 3.3., eq. (14); sec 4.1)
for the DR estimator: non-probability component of the variance for DR estimator
name of the selected link function for the PS model (character)
model type (character)
Łukasz Chrostowski, Maciej Beręsewicz
# Printing information on the model selected
method_ps()
# extracting specific field
method_ps("cloglog")$make_gradient
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.