endo

Study of 79 primarily diagnosed cases of endometrial cancer in order to explain the histology of the endometrium by putative risk factors. The <code>endo</code> data frame has 79 rows and 4 columns. Variable histology, <code>HG</code>. Three risk factors:  variable neovasculization, <code>NV</code>,  variable pulsatility index, <code>PI</code> and variable endometrium height, <code>EH</code>.

datasets

Fit generalized linear models with binomial responses using  a median modified score approach (Kenne Pagui et al., 2016, <https://arxiv.org/abs/1604.04768>) to median bias reduction. This method respects equivariance under reparameterizations for each parameter component and also solves the infinite estimates problem (data separation).

Euloge Clovis Kenne Pagui

mbrglm

Median Bias Reduction in Binomial-Response GLMs

endo function

A data frame with 79 observations on the following 4 variables.
  <dl class="dl-horizontal">
    <dt><code>NV</code></dt><dd>a factor with two levels 1 (present) and 0 (absent) </dd>
    <dt><code>PI</code></dt><dd>pulsatility index of arteria uterina</dd>
    <dt><code>EH</code></dt><dd>endometrium height</dd>
    <dt><code>HG</code></dt><dd>a factor with two levels classified as either 0 (grading 0-II) or 1 (grading III-IV)</dd>
  </dl>

Format


 — endo

A data frame with 79 observations on the following 4 variables.
  <dl class='dl-horizontal'>
    <dt><code>NV</code></dt><dd>a factor with two levels 1 (present) and 0 (absent) </dd>
    <dt><code>PI</code></dt><dd>pulsatility index of arteria uterina</dd>
    <dt><code>EH</code></dt><dd>endometrium height</dd>
    <dt><code>HG</code></dt><dd>a factor with two levels classified as either 0 (grading 0-II) or 1 (grading III-IV)</dd>
  </dl>

endo:

Description

Usage

Arguments

Format

Examples