EVOLUTION-MANAGER
Edit File: augment.mfx.html
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"><html xmlns="http://www.w3.org/1999/xhtml"><head><title>R: Augment data with information from a(n) mfx object</title> <meta http-equiv="Content-Type" content="text/html; charset=utf-8" /> <link rel="stylesheet" type="text/css" href="R.css" /> </head><body> <table width="100%" summary="page for augment.mfx {broom}"><tr><td>augment.mfx {broom}</td><td style="text-align: right;">R Documentation</td></tr></table> <h2>Augment data with information from a(n) mfx object</h2> <h3>Description</h3> <p>Augment accepts a model object and a dataset and adds information about each observation in the dataset. Most commonly, this includes predicted values in the <code>.fitted</code> column, residuals in the <code>.resid</code> column, and standard errors for the fitted values in a <code>.se.fit</code> column. New columns always begin with a <code>.</code> prefix to avoid overwriting columns in the original dataset. </p> <p>Users may pass data to augment via either the <code>data</code> argument or the <code>newdata</code> argument. If the user passes data to the <code>data</code> argument, it <strong>must</strong> be exactly the data that was used to fit the model object. Pass datasets to <code>newdata</code> to augment data that was not used during model fitting. This still requires that all columns used to fit the model are present. </p> <p>Augment will often behave differently depending on whether <code>data</code> or <code>newdata</code> is given. This is because there is often information associated with training observations (such as influences or related) measures that is not meaningfully defined for new observations. </p> <p>For convenience, many augment methods provide default <code>data</code> arguments, so that <code>augment(fit)</code> will return the augmented training data. In these cases, augment tries to reconstruct the original data based on the model object with varying degrees of success. </p> <p>The augmented dataset is always returned as a <a href="../../tibble/html/tibble.html">tibble::tibble</a> with the <strong>same number of rows</strong> as the passed dataset. This means that the passed data must be coercible to a tibble. At this time, tibbles do not support matrix-columns. This means you should not specify a matrix of covariates in a model formula during the original model fitting process, and that <code><a href="../../splines/html/ns.html">splines::ns()</a></code>, <code><a href="../../stats/html/poly.html">stats::poly()</a></code> and <code><a href="../../survival/html/Surv.html">survival::Surv()</a></code> objects are not supported in input data. If you encounter errors, try explicitly passing a tibble, or fitting the original model on data in a tibble. </p> <p>We are in the process of defining behaviors for models fit with various <code>na.action</code> arguments, but make no guarantees about behavior when data is missing at this time. </p> <h3>Usage</h3> <pre> ## S3 method for class 'mfx' augment( x, data = model.frame(x$fit), newdata = NULL, type.predict = c("link", "response", "terms"), type.residuals = c("deviance", "pearson"), se_fit = FALSE, ... ) ## S3 method for class 'logitmfx' augment( x, data = model.frame(x$fit), newdata = NULL, type.predict = c("link", "response", "terms"), type.residuals = c("deviance", "pearson"), se_fit = FALSE, ... ) ## S3 method for class 'negbinmfx' augment( x, data = model.frame(x$fit), newdata = NULL, type.predict = c("link", "response", "terms"), type.residuals = c("deviance", "pearson"), se_fit = FALSE, ... ) ## S3 method for class 'poissonmfx' augment( x, data = model.frame(x$fit), newdata = NULL, type.predict = c("link", "response", "terms"), type.residuals = c("deviance", "pearson"), se_fit = FALSE, ... ) ## S3 method for class 'probitmfx' augment( x, data = model.frame(x$fit), newdata = NULL, type.predict = c("link", "response", "terms"), type.residuals = c("deviance", "pearson"), se_fit = FALSE, ... ) </pre> <h3>Arguments</h3> <table summary="R argblock"> <tr valign="top"><td><code>x</code></td> <td> <p>A <code>logitmfx</code>, <code>negbinmfx</code>, <code>poissonmfx</code>, or <code>probitmfx</code> object. (Note that <code>betamfx</code> objects receive their own set of tidiers.)</p> </td></tr> <tr valign="top"><td><code>data</code></td> <td> <p>A <a href="../../base/html/data.frame.html">base::data.frame</a> or <code><a href="../../tibble/html/tibble.html">tibble::tibble()</a></code> containing the original data that was used to produce the object <code>x</code>. Defaults to <code>stats::model.frame(x)</code> so that <code>augment(my_fit)</code> returns the augmented original data. <strong>Do not</strong> pass new data to the <code>data</code> argument. Augment will report information such as influence and cooks distance for data passed to the <code>data</code> argument. These measures are only defined for the original training data.</p> </td></tr> <tr valign="top"><td><code>newdata</code></td> <td> <p>A <code><a href="../../base/html/data.frame.html">base::data.frame()</a></code> or <code><a href="../../tibble/html/tibble.html">tibble::tibble()</a></code> containing all the original predictors used to create <code>x</code>. Defaults to <code>NULL</code>, indicating that nothing has been passed to <code>newdata</code>. If <code>newdata</code> is specified, the <code>data</code> argument will be ignored.</p> </td></tr> <tr valign="top"><td><code>type.predict</code></td> <td> <p>Passed to <code><a href="../../stats/html/predict.glm.html">stats::predict.glm()</a></code> <code>type</code> argument. Defaults to <code>"link"</code>.</p> </td></tr> <tr valign="top"><td><code>type.residuals</code></td> <td> <p>Passed to <code><a href="../../stats/html/glm.summaries.html">stats::residuals.glm()</a></code> and to <code><a href="../../stats/html/influence.measures.html">stats::rstandard.glm()</a></code> <code>type</code> arguments. Defaults to <code>"deviance"</code>.</p> </td></tr> <tr valign="top"><td><code>se_fit</code></td> <td> <p>Logical indicating whether or not a <code>.se.fit</code> column should be added to the augmented output. For some models, this calculation can be somwhat time-consuming. Defaults to <code>FALSE</code>.</p> </td></tr> <tr valign="top"><td><code>...</code></td> <td> <p>Additional arguments. Not used. Needed to match generic signature only. <strong>Cautionary note:</strong> Misspelled arguments will be absorbed in <code>...</code>, where they will be ignored. If the misspelled argument has a default value, the default value will be used. For example, if you pass <code>conf.lvel = 0.9</code>, all computation will proceed using <code>conf.level = 0.95</code>. Additionally, if you pass <code>newdata = my_tibble</code> to an <code><a href="reexports.html">augment()</a></code> method that does not accept a <code>newdata</code> argument, it will use the default value for the <code>data</code> argument.</p> </td></tr> </table> <h3>Details</h3> <p>This generic augment method wraps <code><a href="augment.glm.html">augment.glm()</a></code> for applicable objects from the <code>mfx</code> package. </p> <h3>Value</h3> <p>A <code><a href="../../tibble/html/tibble.html">tibble::tibble()</a></code> with columns: </p> <table summary="R valueblock"> <tr valign="top"><td><code>.cooksd</code></td> <td> <p>Cooks distance.</p> </td></tr> <tr valign="top"><td><code>.fitted</code></td> <td> <p>Fitted or predicted value.</p> </td></tr> <tr valign="top"><td><code>.hat</code></td> <td> <p>Diagonal of the hat matrix.</p> </td></tr> <tr valign="top"><td><code>.resid</code></td> <td> <p>The difference between observed and fitted values.</p> </td></tr> <tr valign="top"><td><code>.se.fit</code></td> <td> <p>Standard errors of fitted values.</p> </td></tr> <tr valign="top"><td><code>.sigma</code></td> <td> <p>Estimated residual standard deviation when corresponding observation is dropped from model.</p> </td></tr> <tr valign="top"><td><code>.std.resid</code></td> <td> <p>Standardised residuals.</p> </td></tr> </table> <h3>See Also</h3> <p><code><a href="augment.glm.html">augment.glm()</a></code>, <code><a href="../../mfx/html/logitmfx.html">mfx::logitmfx()</a></code>, <code><a href="../../mfx/html/negbinmfx.html">mfx::negbinmfx()</a></code>, <code><a href="../../mfx/html/poissonmfx.html">mfx::poissonmfx()</a></code>, <code><a href="../../mfx/html/probitmfx.html">mfx::probitmfx()</a></code> </p> <p>Other mfx tidiers: <code><a href="augment.betamfx.html">augment.betamfx</a>()</code>, <code><a href="glance.betamfx.html">glance.betamfx</a>()</code>, <code><a href="glance.mfx.html">glance.mfx</a>()</code>, <code><a href="tidy.betamfx.html">tidy.betamfx</a>()</code>, <code><a href="tidy.mfx.html">tidy.mfx</a>()</code> </p> <h3>Examples</h3> <pre> ## Not run: library(mfx) ## Get the marginal effects from a logit regression mod_logmfx <- logitmfx(am ~ cyl + hp + wt, atmean = TRUE, data = mtcars) tidy(mod_logmfx, conf.int = TRUE) ## Compare with the naive model coefficients of the same logit call (not run) # tidy(glm(am ~ cyl + hp + wt, family = binomial, data = mtcars), conf.int = TRUE) augment(mod_logmfx) glance(mod_logmfx) ## Another example, this time using probit regression mod_probmfx <- probitmfx(am ~ cyl + hp + wt, atmean = TRUE, data = mtcars) tidy(mod_probmfx, conf.int = TRUE) augment(mod_probmfx) glance(mod_probmfx) ## End(Not run) </pre> <hr /><div style="text-align: center;">[Package <em>broom</em> version 0.7.0 <a href="00Index.html">Index</a>]</div> </body></html>