myglmmPQL

A flexible version of MASS::glmmPQL

internal

A collection of privacy-preserving distributed algorithms for conducting multi-site data analyses. The regression analyses can be linear regression for continuous outcome, logistic regression for binary outcome, Cox proportional hazard regression for time-to event outcome, Poisson regression for count outcome, or multi-categorical regression for nominal or ordinal outcome. The PDA algorithm runs on a lead site and only requires summary statistics from collaborating sites, with one or few iterations. The package can be used together with the online system (<https://pda-ota.pdamethods.org/>) for safe and convenient collaboration. For more information, please visit our software websites: <https://github.com/Penncil/pda>, and <https://pdamethods.org/>.

Jiajie Chen

Privacy-Preserving Distributed Algorithms

Chongliang Luo

Rui Duan

Mackenzie Edmondson

Jiayi Tong

Xiaokang Liu

Kenneth Locke

Yong Chen

Penn Computing Inference Learning (PennCIL) lab 

myglmmPQL function

<dl><dt>formula.glm</dt>
<dd>formula used to fit <code>glm</code> for initial fixed effects</dd>
<dt>formula</dt>
<dd>formula used to fit iterative <code>lmer</code> in PQL algorithm</dd>
<dt>offset</dt>
<dd><code>glm</code> offset term</dd>
<dt>family</dt>
<dd><code>glm</code> family</dd>
<dt>data</dt>
<dd><code>glm</code> data</dd>
<dt>fixef.init</dt>
<dd>initial fixed effects estimates, set to zeros if NULL</dd>
<dt>weights</dt>
<dd><code>glm</code> weights</dd>
<dt>REML</dt>
<dd><code>lmer</code> logical scalar - Should the estimates be chosen to optimize the REML criterion (as opposed to the log-likelihood)?</dd>
<dt>niter</dt>
<dd><code>glmmPQL</code> maximum number of iterations.</dd>
<dt>verbose</dt>
<dd><code>glmmPQL</code> logical: print out record of iterations?</dd></dl>

Arguments

A flexible version of MASS::glmmPQL — myglmmPQL

<dl>

<dt>formula.glm</dt>
<dd>formula used to fit <code>glm</code> for initial fixed effects</dd>


<dt>formula</dt>
<dd>formula used to fit iterative <code>lmer</code> in PQL algorithm</dd>


<dt>offset</dt>
<dd><code>glm</code> offset term</dd>


<dt>family</dt>
<dd><code>glm</code> family</dd>


<dt>data</dt>
<dd><code>glm</code> data</dd>


<dt>fixef.init</dt>
<dd>initial fixed effects estimates, set to zeros if NULL</dd>


<dt>weights</dt>
<dd><code>glm</code> weights</dd>


<dt>REML</dt>
<dd><code>lmer</code> logical scalar - Should the estimates be chosen to optimize the REML criterion (as opposed to the log-likelihood)?</dd>


<dt>niter</dt>
<dd><code>glmmPQL</code> maximum number of iterations.</dd>


<dt>verbose</dt>
<dd><code>glmmPQL</code> logical: print out record of iterations?</dd>

</dl>

A flexible version of MASS::glmmPQL

myglmmPQL: A flexible version of MASS::glmmPQL

Description

Usage

Value

Arguments

Details