score

Statistical modeling for correlated count data using the beta-binomial distribution, described in Martin et al. (2020) <doi:10.1214/19-AOAS1283>. It allows for both mean and overdispersion covariates.

Amy D Willis

corncob

Count Regression for Correlated Observations with the
Beta-Binomial

Bryan D Martin

Daniela Witten

Sarah Teichman

Thomas W Yee

Xiangjie Xue

score function

<dl><dt>mod</dt>
<dd>an object of class <code>bbdml</code></dd>
<dt>numerical</dt>
<dd>Boolean. Defaults to <code>FALSE</code>. Indicator of whether to use the numeric Hessian and score (not recommended).</dd>
<dt>get_score_covariance</dt>
<dd>Boolean. Defaults to <code>FALSE</code>. Should we return a robust estimate of variance of score: \(\hat{B}(\hat{\theta}) = \sum_i G(\hat{\theta}; W_i) G(\hat{\theta}; W_i)^T\). This parameter is not intended for users.</dd></dl>

Arguments

Compute score at the MLE — score

<dl>

<dt>mod</dt>
<dd>an object of class <code>bbdml</code></dd>


<dt>numerical</dt>
<dd>Boolean. Defaults to <code>FALSE</code>. Indicator of whether to use the numeric Hessian and score (not recommended).</dd>


<dt>get_score_covariance</dt>
<dd>Boolean. Defaults to <code>FALSE</code>. Should we return a robust estimate of variance of score: \(\hat{B}(\hat{\theta}) = \sum_i G(\hat{\theta}; W_i) G(\hat{\theta}; W_i)^T\). This parameter is not intended for users.</dd>

</dl>

Compute score at the MLE

Last chance! 50% off unlimited learning

score: Compute score at the MLE

Description

Usage

Value

Arguments

Examples