optimal_g

Computes the value of g such that width of the \(1-\alpha\) confidence interval
at sample size n is minimized

Anytime-valid inference for linear models, namely, sequential t-tests, sequential F-tests, and confidence sequences with time-uniform Type-I error and coverage guarantees. This allows hypotheses to be continuously tested without sacrificing false positive guarantees. It is based on the methods documented in Lindon et al. (2022) <doi:10.48550/arXiv.2210.08589>.

Michael Lindon

avlm

Safe Anytime Valid Inference for Linear Models

optimal_g function

<dl><dt>n</dt>
<dd>A positive sample size integer.</dd>
<dt>number_of_coefficients</dt>
<dd>A positive integer of coefficients in the full model</dd>
<dt>alpha</dt>
<dd>A positive numeric scalar in (0,1) for nominal Type I error.</dd></dl>

Arguments

Computes the value of g such that width of the \(1-\alpha\) confidence interval
at sample size n is minimized — optimal_g

<dl>

<dt>n</dt>
<dd>A positive sample size integer.</dd>


<dt>number_of_coefficients</dt>
<dd>A positive integer of coefficients in the full model</dd>


<dt>alpha</dt>
<dd>A positive numeric scalar in (0,1) for nominal Type I error.</dd>

</dl>

optimal_g: Computes the value of g such that width of the \(1-\alpha\) confidence interval at sample size n is minimized

Description

Usage

Value

Arguments

Examples