More generally, to create a linear regression model, use the fitlm()
function. This function creates a LinearModel
object. The object in the workspace has a series of properties that can be immediately viewed by simply clicking on it. Methods such as plot
, plotResiduals
, and plotDiagnostics
are available if you want to create plots and perform a diagnostic analysis.
Note
LinearModel
is an object comprising training data, model description, diagnostic information, and fitted coefficients for a linear regression.
By default, fitlm()
takes the last variable in the table or dataset array as the response. Otherwise, we have to specify predictors and response variables, for example, as a formula. In addition, we can set a specific column as the response variable by using the ResponseVar
name-value pair argument. To use a set of the columns as predictors, use the PredictorVars
name-value pair argument. Predictor variables can be numeric or of any grouping variable...