predict.FGLMtrunc

This function returns truncated estimate of linear predictors, fitted values, and functional parameter \(\beta\)
for a fitted <code>FGLMtrunc</code> object.

An implementation of the methodologies described in Xi Liu, Afshin A. Divani, and Alexander Petersen (2022) <doi:10.1016/j.csda.2022.107421>, including
truncated functional linear and truncated functional logistic regression models.

Chau Tran

FGLMtrunc

Truncated Functional Generalized Linear Models

Xi Liu

Alexander Petersen

predict.FGLMtrunc function

<dl><dt>object</dt>
<dd>fitted <code>FGLMtrunc</code> object</dd>
<dt>newX.curves</dt>
<dd>Matrix of new values for functional predictors <code>X.curves</code>.</dd>
<dt>newS</dt>
<dd>Matrix of new values for scalar predictors <code>S</code>.</dd>
<dt>type</dt>
<dd>Type of prediction. For logistic regression (<code>family = "binomial"</code>), <code>type="link"</code> gives the linear
predictors, which is log-odds, and <code>type="response"</code> gives the predicted probabilities.
For linear regression (<code>family = "gaussian"</code>), both <code>type="link"</code> and <code>type="response"</code> give fitted values.
For both linear regression and logistic regression, <code>type="coefficients"</code> gives truncated estimate of functional parameter \(\beta\).</dd>
<dt>...</dt>
<dd>additional predict arguments (Not applicable for FGLMtrunc)</dd></dl>

Arguments

Make predictions from FGLMtrunc fitted model — predict.FGLMtrunc

<dl>

<dt>object</dt>
<dd>fitted <code>FGLMtrunc</code> object</dd>


<dt>newX.curves</dt>
<dd>Matrix of new values for functional predictors <code>X.curves</code>.</dd>


<dt>newS</dt>
<dd>Matrix of new values for scalar predictors <code>S</code>.</dd>


<dt>type</dt>
<dd>Type of prediction. For logistic regression (<code>family = "binomial"</code>), <code>type="link"</code> gives the linear
predictors, which is log-odds, and <code>type="response"</code> gives the predicted probabilities.
For linear regression (<code>family = "gaussian"</code>), both <code>type="link"</code> and <code>type="response"</code> give fitted values.
For both linear regression and logistic regression, <code>type="coefficients"</code> gives truncated estimate of functional parameter \(\beta\).</dd>


<dt>...</dt>
<dd>additional predict arguments (Not applicable for FGLMtrunc)</dd>

</dl>

predict.FGLMtrunc: Make predictions from `FGLMtrunc` fitted model

Description

Usage

Value

Arguments

See Also