Regression - Save Variable(s) - Predicted Values

From Q
Jump to: navigation, search

Creates a new variable containing predicted values for each case in the data from a regression. Observations with missing values in the predictors are not predicted.


Running this QScript will produce a new variable, which will appear in your data set for use in further analyses. For a categorical outcome (Regression - Ordered Logit, Regression - Binary Logit, Regression - Multinomial Logit) the variable will be categorical containing the predicted category for each observation.

Technical details

Uses the predict method from R. Unlike the defaults in most R packages, this implementation assigns predicted values for observations not in the estimation sample. Categorical variables are predicted as discrete (e.g., in Regression - Binary Logit, the predicted values are of the two categories). See Regression - Save Variable(s) - Fitted Values for non-discrete value predictions.


includeWeb("QScript R Output Functions");
function main() {
    saveVariables("Predicted values", "Regression", "predict(", ")", null, null, "pred.val");