fast_forward

Add variables to a model as long as they reduce the given criterion.
Variables are searched according to <code>candidates</code> and every one which
reduces the criterion is added (not necessarily the best one).

Selecting linear and generalized linear models for large data sets
using modified stepwise procedure and modern selection criteria (like
modifications of Bayesian Information Criterion). Selection can be
performed on data which exceed RAM capacity. Bogdan et al., (2004)
<doi:10.1534/genetics.103.021683>.

Piotr Szulc

bigstep

Stepwise Selection for Large Data Sets

fast_forward function

<dl><dt>data</dt>
<dd>an object of class <code>big</code>.</dd>
<dt>crit</dt>
<dd>a function defining the model selection criterion. You can use
your own function or one of these: <code>bic</code>, <code>mbic</code>, <code>mbic2</code>,
<code>aic</code>, <code>maic</code>, <code>maic2</code>.</dd>
<dt>...</dt>
<dd>optional arguments to <code>crit</code>.</dd>
<dt>maxf</dt>
<dd>a numeric, a maximal number of variables in the final model.</dd></dl>

Arguments

Fast-forward step — fast_forward

<dl>

<dt>data</dt>
<dd>an object of class <code>big</code>.</dd>


<dt>crit</dt>
<dd>a function defining the model selection criterion. You can use
your own function or one of these: <code>bic</code>, <code>mbic</code>, <code>mbic2</code>,
<code>aic</code>, <code>maic</code>, <code>maic2</code>.</dd>


<dt>...</dt>
<dd>optional arguments to <code>crit</code>.</dd>


<dt>maxf</dt>
<dd>a numeric, a maximal number of variables in the final model.</dd>

</dl>

fast_forward: Fast-forward step

Description

Usage

Value

Arguments

Details

Examples