LinearRateChangeDrop

Creates an MAR <code>DropoutMechanism</code> object where subject <code>i</code> has piecewise exponential
dropout rate where the rate changes by a constant amount after each event, specifically after <code>j</code> events
the subject has rate <code>Rij = Cj*exp(Xij)</code> where <code>Cj=C+j*D</code> for constants <code>C</code>, <code>D</code> 
and Xij is a standard normal variable with mean 0 and standard deviation <code>sigma</code>

Performs reference based multiple imputation of recurrent event data
based on a negative binomial regression model, as described
by Keene et al (2014) <doi:10.1002/pst.1624>.

Jonathan Bartlett

dejaVu

Multiple Imputation for Recurrent Events

Nikolas Burkoff

Paul Metcalfe

David Ruau

LinearRateChangeDrop function

<dl><dt>starting.rate</dt>
<dd><code>C</code>, see description section.</dd>
<dt>rate.change</dt>
<dd><code>D</code>, see description section. Note if <code>D&lt;0</code>, <code>Cj</code> could be negative
for large <code>j</code>, this is not possible and the rate remains constant if the next change would set <code>Cj&lt;=0</code></dd>
<dt>var</dt>
<dd><code>sigma^2</code>, see description section</dd></dl>

Arguments

Create a Dropout Mechanism with drop out rate which changes by a fixed constant
after every event — LinearRateChangeDrop

<dl>

<dt>starting.rate</dt>
<dd><code>C</code>, see description section.</dd>


<dt>rate.change</dt>
<dd><code>D</code>, see description section. Note if <code>D&lt;0</code>, <code>Cj</code> could be negative
for large <code>j</code>, this is not possible and the rate remains constant if the next change would set <code>Cj&lt;=0</code></dd>


<dt>var</dt>
<dd><code>sigma^2</code>, see description section</dd>

</dl>

LinearRateChangeDrop: Create a Dropout Mechanism with drop out rate which changes by a fixed constant after every event

Description

Usage

Value

Arguments

See Also

Examples