Ordinary least squares regression.
Usage
ols_regress(object, ...)
# S3 method for class 'lm'
ols_regress(object, conf.level = 0.95, iterm = FALSE, title = "model", ...)Arguments
- object
An object of class "formula" (or one that can be coerced to that class): a symbolic description of the model to be fitted or class
lm.- ...
Other inputs.
- conf.level
Confidence level for the confidence intervals of the coefficients.
- iterm
Logical; if
TRUE, the predictors are scaled and centered before computing the standardized betas. See Interaction Terms below.- title
Character; title of the model, printed as the header of the parameter estimates table.
Value
ols_regress returns an object of class "ols_regress".
An object of class "ols_regress" is a list containing the following
components:
- r
square root of rsquare, correlation between observed and predicted values of dependent variable
- rsq
coefficient of determination or r-square
- adjr
adjusted rsquare
- rmse
root mean squared error
- cv
coefficient of variation
- mse
mean squared error
- mae
mean absolute error
- aic
akaike information criteria
- sbc
bayesian information criteria
- sbic
sawa bayesian information criteria
- prsq
predicted rsquare
- error_df
residual degrees of freedom
- model_df
regression degrees of freedom
- total_df
total degrees of freedom
- ess
error sum of squares
- rss
regression sum of squares
- tss
total sum of squares
- rms
regression mean square
- ems
error mean square
- f
f statistics
- p
p-value for
f- n
number of predictors including intercept
- betas
betas; estimated coefficients
- sbetas
standardized betas
- std_errors
standard errors
- tvalues
t values
- pvalues
p-value of
tvalues- df
degrees of freedom of
betas- conf_lm
confidence intervals for coefficients
- title
title for the model
- dependent
character vector; name of the dependent variable
- predictors
character vector; name of the predictor variables
- mvars
character vector; name of the predictor variables including intercept
- model
input model for
ols_regress
Details
Refits accept explicit contrast matrices or the base contrast names
contr.treatment, contr.sum, contr.helmert, contr.poly and contr.SAS
(resolved from the stats namespace); models fitted with custom named contrast
functions must be refitted with an explicit matrix, for example
contrasts = list(g = my_matrix). Refits carry the fitted model's frozen total
offset as an argument with all formula offset() terms removed for
standardized-beta and selection computations. Returned stepwise models are
fitted from the original call, preserving formula and argument offsets for
prediction on new data. Their coefficients, fitted values and total offset
must be identical to the frozen-offset selection fit; otherwise selection
stops with an error because prediction with new data is not supported for
that model.
For non-hierarchical stepwise selection, candidate rankings and criterion improvement tests use 10 significant digits: values that agree to 10 significant digits are ties, equal criterion values mean no improvement, and a remaining tie goes to the first candidate in formula order after any F tie is broken by p value (smaller for entry, larger for removal). P-value entry and removal cutoffs are compared without rounding.
Interaction Terms
If the model includes interaction terms, the standardized betas are computed after scaling and centering the predictors (with the model's prior weights when the model is weighted) and refitting the model.
Weighted Models
Models fitted with lm(..., weights = ...) are supported. The reported
estimates, standard errors, test statistics, confidence intervals and
R-squared match summary(), anova() and confint().
Weights cannot be supplied through the formula interface; fit the model with
lm() and pass the fitted model instead. Objects of class glm
are not supported; use lm() or the formula interface.
References
https://www.ssc.wisc.edu/~hemken/Stataworkshops/stdBeta/Getting%20Standardized%20Coefficients%20Right.pdf
Kutner, M. H., Nachtsheim, C. J., Neter, J. and Li, W. (2005). Applied Linear Statistical Models, 5th ed. New York: McGraw-Hill/Irwin.
Aiken, L. S. and West, S. G. (1991). Multiple Regression: Testing and Interpreting Interactions. Newbury Park, CA: Sage.
Examples
ols_regress(mpg ~ disp + hp + wt, data = mtcars)
#> Model Summary
#> ---------------------------------------------------------------
#> R 0.909 RMSE 2.639
#> R-Squared 0.827 MSE 6.964
#> Adj. R-Squared 0.808 Coef. Var 13.135
#> Pred R-Squared 0.768 AIC 158.643
#> MAE 1.907 SBC 165.972
#> ---------------------------------------------------------------
#> RMSE: Root Mean Square Error
#> MSE: Mean Square Error
#> MAE: Mean Absolute Error
#> AIC: Akaike Information Criteria
#> SBC: Schwarz Bayesian Criteria
#>
#> ANOVA
#> --------------------------------------------------------------------
#> Sum of
#> Squares DF Mean Square F Sig.
#> --------------------------------------------------------------------
#> Regression 931.057 3 310.352 44.566 0.0000
#> Residual 194.991 28 6.964
#> Total 1126.047 31
#> --------------------------------------------------------------------
#>
#> Parameter Estimates
#> ----------------------------------------------------------------------------------------
#> model Beta Std. Error Std. Beta t Sig lower upper
#> ----------------------------------------------------------------------------------------
#> (Intercept) 37.106 2.111 17.579 0.000 32.782 41.429
#> disp -0.001 0.010 -0.019 -0.091 0.929 -0.022 0.020
#> hp -0.031 0.011 -0.354 -2.724 0.011 -0.055 -0.008
#> wt -3.801 1.066 -0.617 -3.565 0.001 -5.985 -1.617
#> ----------------------------------------------------------------------------------------
# if model includes interaction terms set iterm to TRUE
ols_regress(mpg ~ disp * wt, data = mtcars, iterm = TRUE)
#> Model Summary
#> ---------------------------------------------------------------
#> R 0.922 RMSE 2.455
#> R-Squared 0.850 MSE 6.027
#> Adj. R-Squared 0.834 Coef. Var 12.219
#> Pred R-Squared 0.801 AIC 154.018
#> MAE 1.897 SBC 161.346
#> ---------------------------------------------------------------
#> RMSE: Root Mean Square Error
#> MSE: Mean Square Error
#> MAE: Mean Absolute Error
#> AIC: Akaike Information Criteria
#> SBC: Schwarz Bayesian Criteria
#>
#> ANOVA
#> --------------------------------------------------------------------
#> Sum of
#> Squares DF Mean Square F Sig.
#> --------------------------------------------------------------------
#> Regression 957.299 3 319.100 52.947 0.0000
#> Residual 168.749 28 6.027
#> Total 1126.047 31
#> --------------------------------------------------------------------
#>
#> Parameter Estimates
#> ----------------------------------------------------------------------------------------
#> model Beta Std. Error Std. Beta t Sig lower upper
#> ----------------------------------------------------------------------------------------
#> (Intercept) 44.082 3.123 14.115 0.000 37.685 50.479
#> disp -0.056 0.013 -0.385 -4.257 0.000 -0.083 -0.029
#> wt -6.496 1.313 -0.616 -4.946 0.000 -9.186 -3.805
#> disp:wt 0.012 0.003 0.278 3.596 0.001 0.005 0.018
#> ----------------------------------------------------------------------------------------