Wald_trans.Wald_robust {cta} | R Documentation |
Wald-Type CIs (Robust)
Description
Constructs non-transformed and transformed (if the
transformation g
is specified) Wald confidence intervals (CIs) for estimands in contingency tables subject to equality constraints.
Usage
Wald_trans.Wald_robust(y, strata, fixed.strata, h0.fct, h0.fct.deriv,
S0.fct, S0.fct.deriv, max.mph.iter, step,
change.step.after, y.eps, iter.orig, norm.diff.conv,
norm.score.conv, max.score.diff.iter, cut.off,
S.space.H0, trans.g, trans.g.deriv, trans.g.inv,
adj.epsilon, iter.robust.max, iter.robust.eff)
Arguments
y |
Observed table counts in the contingency table(s), in vector form. |
strata |
Vector of the same length as |
fixed.strata |
The object that gives information on which stratum (strata) has (have) fixed sample sizes. |
h0.fct |
The constraint function |
h0.fct.deriv |
The R function object that computes analytic derivative of the
transpose of the constraint function |
S0.fct |
The estimand function |
S0.fct.deriv |
The R function object that computes analytic derivative of
the estimand function |
max.mph.iter , step , change.step.after , y.eps , iter.orig , norm.diff.conv , norm.score.conv , max.score.diff.iter |
The parameters used in |
cut.off |
|
S.space.H0 |
Restricted estimand space of |
trans.g |
The transformation |
trans.g.deriv |
The derivative function of the transformation |
trans.g.inv |
|
adj.epsilon , iter.robust.max , iter.robust.eff |
The parameters used in the robustifying procedure. |
Value
Wald_trans.Wald_robust
returns a list, which includes two objects. The first object is
either a
1
-by-2
matrix which displays two endpoints of the non-transformed Wald confidence interval, if the transformationg
is not specified;or a
2
-by-2
matrix, whose first row displays two endpoints of the non-transformed Wald confidence interval, and whose second row displays two endpoints of the transformed Wald confidence interval, if the transformationg
is specified.
For the second object, it includes the warning message that occurs during construction of the confidence
interval(s) if the robustifying procedure is evoked: "Wald.CI: Adjustment used. Not on original data.\n"
, or "Wald.CI and trans.Wald.CI: Adjustment used. Not on original data.\n"
. If the robustifying procedure is not evoked, the second object is NULL
.
Author(s)
Qiansheng Zhu
References
Zhu, Q. (2020) "On improved confidence intervals for parameters of discrete distributions." PhD dissertation, University of Iowa.
See Also
Wald_trans.Wald_nr
, f.psi
, ci.table