check_and_align_vars

Ensures that variables in <code>grid_df</code> are type-compatible with <code>df_fit</code> and that
factor (including ordered factor) levels are aligned to those used during model
fitting. Character columns in <code>df_fit</code> are first coerced to factors. For any
factor/ordered variable, <code>grid_df</code> is coerced to the same type and levels; any
unseen levels in <code>grid_df</code> will trigger an error.

internal

Provides computational tools for estimating inverse regions and
constructing the corresponding simultaneous outer and inner confidence regions.
Acceptable input includes both one-dimensional and two-dimensional data for linear,
logistic, functional, and spatial generalized least squares regression models.
Functions are also available for constructing simultaneous confidence bands (SCBs)
for these models. The definition of simultaneous confidence regions (SCRs) follows
Sommerfeld et al. (2018) <doi:10.1080/01621459.2017.1341838>. Methods for estimating
inverse regions, SCRs, and the nonparametric bootstrap are based on
Ren et al. (2024) <doi:10.1093/jrsssc/qlae027>. Methods for constructing SCBs
are described in Crainiceanu et al. (2024) <doi:10.1201/9781003278726>
and Telschow et al. (2022) <doi:10.1016/j.jspi.2021.05.008>.

Zhuoran Yu

SCoRES

Simultaneous Confidence Region Excursion Sets

Armin Schwartzman

Junting Ren

Julia Wrobel

check_and_align_vars function

<dl><dt>df_fit</dt>
<dd>A data frame used for model fitting. Character columns will be
coerced to factors before alignment.</dd>
<dt>grid_df</dt>
<dd>A data frame of covariate settings (newdata) at which
predictions/SCBs are to be evaluated.</dd>
<dt>model_vars</dt>
<dd>A vector contained all the interested columns that appear
in <code>df_fit</code>. Only those variables are aligned in <code>grid_df</code>;
other columns are left unchanged. Default is NULL.</dd>
<dt>grid_boot</dt>
<dd>A logic value for whether the function is for constructing
<code>grid_df_boot</code> or not.</dd></dl>

Arguments

Validate and align types/levels between training data and prediction grid — check_and_align_vars

<dl>

<dt>df_fit</dt>
<dd>A data frame used for model fitting. Character columns will be
coerced to factors before alignment.</dd>


<dt>grid_df</dt>
<dd>A data frame of covariate settings (newdata) at which
predictions/SCBs are to be evaluated.</dd>


<dt>model_vars</dt>
<dd>A vector contained all the interested columns that appear
in <code>df_fit</code>. Only those variables are aligned in <code>grid_df</code>;
other columns are left unchanged. Default is NULL.</dd>


<dt>grid_boot</dt>
<dd>A logic value for whether the function is for constructing
<code>grid_df_boot</code> or not.</dd>

</dl>

check_and_align_vars: Validate and align types/levels between training data and prediction grid

Description

Usage

Value

Arguments

Examples