Описание тега mlm

"mlm" model class is not widely known by lm() users. For an introductory example, please refer to Fitting a linear model with multiple LHS. Also read Run lm with multiple responses and weights for the limitations of "mlm".

R does not have full support for "mlm" class at the moment. Here is an overview:

• The following generic functions work and would return a matrix:

  coef(), residuals(), fitted(), vcov()
  

• The following generic functions work and would return a list:

  summary(), anova()
  

• The following generic functions have limited support for "mlm":

  predict()
  

• The following generic functions do not work at all for "mlm":

  rstandard(), plot(), confint()
  

The aim of this wiki page is to direct "mlm" users to efficient solutions of various inference and diagnostic tasks. Here is a good collection:

Model Summary

• Obtain standard errors of regression coefficients for an “mlm” object returned by lm()
• Obtain t-statistic for regression coefficients of an “mlm” object returned by lm()
• Get confidence intervals for regression coefficients of “mlm” object returned by lm()
• Obtain residual standard errors of an “mlm” object returned by lm()

Model Prediction

• Prediction of 'mlm' linear model object from lm()
• Applying lm() and predict() to multiple columns in a data frame

Model Diagnostic

• Obtain standardised residuals and “Residual v.s. Fitted” plot for “mlm” object from lm()

Please read above questions carefully, and only ask a new question when it is not answered in above.