| Title: | Tools for Building OLS Regression Models |
|---|---|
| Description: | Tools designed to make it easier for users, particularly beginner/intermediate R users to build ordinary least squares regression models. Includes comprehensive regression output, heteroskedasticity tests, collinearity diagnostics, residual diagnostics, measures of influence, model fit assessment and variable selection procedures. |
| Authors: | Aravind Hebbali [aut, cre] |
| Maintainer: | Aravind Hebbali <[email protected]> |
| License: | MIT + file LICENSE |
| Version: | 0.6.1 |
| Built: | 2024-11-07 13:45:35 UTC |
| Source: | CRAN |
Akaike information criterion for model selection.
ols_aic(model, method = c("R", "STATA", "SAS"), corrected = FALSE)ols_aic(model, method = c("R", "STATA", "SAS"), corrected = FALSE)
model |
An object of class |
method |
A character vector; specify the method to compute AIC. Valid options include R, STATA and SAS. |
corrected |
Logical; if |
AIC provides a means for model selection. Given a collection of models for the data, AIC estimates the quality of each model, relative to each of the other models. R and STATA use loglikelihood to compute AIC. SAS uses residual sum of squares. Below is the formula in each case:
R & STATA
SAS
corrected
where n is the sample size and p is the number of model parameters including intercept.
Akaike information criterion of the model.
Akaike, H. (1969). “Fitting Autoregressive Models for Prediction.” Annals of the Institute of Statistical Mathematics 21:243–247.
Judge, G. G., Griffiths, W. E., Hill, R. C., and Lee, T.-C. (1980). The Theory and Practice of Econometrics. New York: John Wiley & Sons.
Other model selection criteria:
ols_apc(),
ols_fpe(),
ols_hsp(),
ols_mallows_cp(),
ols_msep(),
ols_sbc(),
ols_sbic()
# using R computation method model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_aic(model) # using STATA computation method model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_aic(model, method = 'STATA') # using SAS computation method model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_aic(model, method = 'SAS') # corrected akaike information criterion model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_aic(model, method = 'SAS', corrected = TRUE)# using R computation method model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_aic(model) # using STATA computation method model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_aic(model, method = 'STATA') # using SAS computation method model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_aic(model, method = 'SAS') # corrected akaike information criterion model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_aic(model, method = 'SAS', corrected = TRUE)
Amemiya's prediction error.
ols_apc(model)ols_apc(model)
model |
An object of class |
Amemiya's Prediction Criterion penalizes R-squared more heavily than does adjusted R-squared for each addition degree of freedom used on the right-hand-side of the equation. The lower the better for this criterion.
where n is the sample size, p is the number of predictors including the intercept and R^2 is the coefficient of determination.
Amemiya's prediction error of the model.
Amemiya, T. (1976). Selection of Regressors. Technical Report 225, Stanford University, Stanford, CA.
Judge, G. G., Griffiths, W. E., Hill, R. C., and Lee, T.-C. (1980). The Theory and Practice of Econometrics. New York: John Wiley & Sons.
Other model selection criteria:
ols_aic(),
ols_fpe(),
ols_hsp(),
ols_mallows_cp(),
ols_msep(),
ols_sbc(),
ols_sbic()
model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_apc(model)model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_apc(model)
Variance inflation factor, tolerance, eigenvalues and condition indices.
ols_coll_diag(model) ols_vif_tol(model) ols_eigen_cindex(model)ols_coll_diag(model) ols_vif_tol(model) ols_eigen_cindex(model)
model |
An object of class |
Collinearity implies two variables are near perfect linear combinations of one another. Multicollinearity involves more than two variables. In the presence of multicollinearity, regression estimates are unstable and have high standard errors.
Tolerance
Percent of variance in the predictor that cannot be accounted for by other predictors.
Steps to calculate tolerance:
Regress the kth predictor on rest of the predictors in the model.
Compute - the coefficient of determination from the regression in the above step.
Variance Inflation Factor
Variance inflation factors measure the inflation in the variances of the parameter estimates due to
collinearities that exist among the predictors. It is a measure of how much the variance of the estimated
regression coefficient is inflated by the existence of correlation among the predictor variables
in the model. A VIF of 1 means that there is no correlation among the kth predictor and the remaining predictor
variables, and hence the variance of is not inflated at all. The general rule of thumb is that VIFs
exceeding 4 warrant further investigation, while VIFs exceeding 10 are signs of serious multicollinearity
requiring correction.
Steps to calculate VIF:
Regress the kth predictor on rest of the predictors in the model.
Compute - the coefficient of determination from the regression in the above step.
Condition Index
Most multivariate statistical approaches involve decomposing a correlation matrix into linear combinations of variables. The linear combinations are chosen so that the first combination has the largest possible variance (subject to some restrictions), the second combination has the next largest variance, subject to being uncorrelated with the first, the third has the largest possible variance, subject to being uncorrelated with the first and second, and so forth. The variance of each of these linear combinations is called an eigenvalue. Collinearity is spotted by finding 2 or more variables that have large proportions of variance (.50 or more) that correspond to large condition indices. A rule of thumb is to label as large those condition indices in the range of 30 or larger.
ols_coll_diag returns an object of class "ols_coll_diag".
An object of class "ols_coll_diag" is a list containing the
following components:
vif_t |
tolerance and variance inflation factors |
eig_cindex |
eigen values and condition index |
Belsley, D. A., Kuh, E., and Welsch, R. E. (1980). Regression Diagnostics: Identifying Influential Data and Sources of Collinearity. New York: John Wiley & Sons.
# model model <- lm(mpg ~ disp + hp + wt + drat, data = mtcars) # vif and tolerance ols_vif_tol(model) # eigenvalues and condition indices ols_eigen_cindex(model) # collinearity diagnostics ols_coll_diag(model)# model model <- lm(mpg ~ disp + hp + wt + drat, data = mtcars) # vif and tolerance ols_vif_tol(model) # eigenvalues and condition indices ols_eigen_cindex(model) # collinearity diagnostics ols_coll_diag(model)
Zero-order, part and partial correlations.
ols_correlations(model)ols_correlations(model)
model |
An object of class |
ols_correlations() returns the relative importance of independent
variables in determining response variable. How much each variable uniquely
contributes to rsquare over and above that which can be accounted for by the
other predictors? Zero order correlation is the Pearson correlation
coefficient between the dependent variable and the independent variables.
Part correlations indicates how much rsquare will decrease if that variable
is removed from the model and partial correlations indicates amount of
variance in response variable, which is not estimated by the other
independent variables in the model, but is estimated by the specific
variable.
ols_correlations returns an object of class "ols_correlations".
An object of class "ols_correlations" is a data frame containing the
following components:
Zero-order |
zero order correlations |
Partial |
partial correlations |
Part |
part correlations |
Morrison, D. F. 1976. Multivariate statistical methods. New York: McGraw-Hill.
model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_correlations(model)model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_correlations(model)
Estimated mean square error of prediction.
ols_fpe(model)ols_fpe(model)
model |
An object of class |
Computes the estimated mean square error of prediction for each model selected assuming that the values of the regressors are fixed and that the model is correct.
where , n is the sample size and p is the number of predictors including the intercept
Final prediction error of the model.
Akaike, H. (1969). “Fitting Autoregressive Models for Prediction.” Annals of the Institute of Statistical Mathematics 21:243–247.
Judge, G. G., Griffiths, W. E., Hill, R. C., and Lee, T.-C. (1980). The Theory and Practice of Econometrics. New York: John Wiley & Sons.
Other model selection criteria:
ols_aic(),
ols_apc(),
ols_hsp(),
ols_mallows_cp(),
ols_msep(),
ols_sbc(),
ols_sbic()
model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_fpe(model)model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_fpe(model)
Measure of influence based on the fact that influential observations in either the response variable or in the predictors or both.
ols_hadi(model)ols_hadi(model)
model |
An object of class |
Hadi's measure of the model.
Chatterjee, Samprit and Hadi, Ali. Regression Analysis by Example. 5th ed. N.p.: John Wiley & Sons, 2012. Print.
Other influence measures:
ols_leverage(),
ols_pred_rsq(),
ols_press()
model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_hadi(model)model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_hadi(model)
Average prediction mean squared error.
ols_hsp(model)ols_hsp(model)
model |
An object of class |
Hocking's Sp criterion is an adjustment of the residual sum of Squares. Minimize this criterion.
where , n is the sample size and p is the number of predictors including the intercept
Hocking's Sp of the model.
Hocking, R. R. (1976). “The Analysis and Selection of Variables in a Linear Regression.” Biometrics 32:1–50.
Other model selection criteria:
ols_aic(),
ols_apc(),
ols_fpe(),
ols_mallows_cp(),
ols_msep(),
ols_sbc(),
ols_sbic()
model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_hsp(model)model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_hsp(model)
Launches shiny app for interactive model building.
ols_launch_app()ols_launch_app()
## Not run: ols_launch_app() ## End(Not run)## Not run: ols_launch_app() ## End(Not run)
The leverage of an observation is based on how much the observation's value on the predictor variable differs from the mean of the predictor variable. The greater an observation's leverage, the more potential it has to be an influential observation.
ols_leverage(model)ols_leverage(model)
model |
An object of class |
Leverage of the model.
Kutner, MH, Nachtscheim CJ, Neter J and Li W., 2004, Applied Linear Statistical Models (5th edition). Chicago, IL., McGraw Hill/Irwin.
Other influence measures:
ols_hadi(),
ols_pred_rsq(),
ols_press()
model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_leverage(model)model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_leverage(model)
Mallow's Cp.
ols_mallows_cp(model, fullmodel)ols_mallows_cp(model, fullmodel)
model |
An object of class |
fullmodel |
An object of class |
Mallows' Cp statistic estimates the size of the bias that is introduced into the predicted responses by having an underspecified model. Use Mallows' Cp to choose between multiple regression models. Look for models where Mallows' Cp is small and close to the number of predictors in the model plus the constant (p).
Mallow's Cp of the model.
Hocking, R. R. (1976). “The Analysis and Selection of Variables in a Linear Regression.” Biometrics 32:1–50.
Mallows, C. L. (1973). “Some Comments on Cp.” Technometrics 15:661–675.
Other model selection criteria:
ols_aic(),
ols_apc(),
ols_fpe(),
ols_hsp(),
ols_msep(),
ols_sbc(),
ols_sbic()
full_model <- lm(mpg ~ ., data = mtcars) model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_mallows_cp(model, full_model)full_model <- lm(mpg ~ ., data = mtcars) model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_mallows_cp(model, full_model)
Estimated error of prediction, assuming multivariate normality.
ols_msep(model)ols_msep(model)
model |
An object of class |
Computes the estimated mean square error of prediction assuming that both independent and dependent variables are multivariate normal.
where , n is the sample size and p is the number of
predictors including the intercept
Estimated error of prediction of the model.
Stein, C. (1960). “Multiple Regression.” In Contributions to Probability and Statistics: Essays in Honor of Harold Hotelling, edited by I. Olkin, S. G. Ghurye, W. Hoeffding, W. G. Madow, and H. B. Mann, 264–305. Stanford, CA: Stanford University Press.
Darlington, R. B. (1968). “Multiple Regression in Psychological Research and Practice.” Psychological Bulletin 69:161–182.
Other model selection criteria:
ols_aic(),
ols_apc(),
ols_fpe(),
ols_hsp(),
ols_mallows_cp(),
ols_sbc(),
ols_sbic()
model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_msep(model)model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_msep(model)
Added variable plot provides information about the marginal importance of a predictor variable, given the other predictor variables already in the model. It shows the marginal importance of the variable in reducing the residual variability.
ols_plot_added_variable(model, print_plot = TRUE)ols_plot_added_variable(model, print_plot = TRUE)
model |
An object of class |
print_plot |
logical; if |
The added variable plot was introduced by Mosteller and Tukey (1977). It enables us to visualize the regression coefficient of a new variable being considered to be included in a model. The plot can be constructed for each predictor variable.
Let us assume we want to test the effect of adding/removing variable X from a model. Let the response variable of the model be Y
Steps to construct an added variable plot:
Regress Y on all variables other than X and store the residuals (Y residuals).
Regress X on all the other variables included in the model (X residuals).
Construct a scatter plot of Y residuals and X residuals.
What do the Y and X residuals represent? The Y residuals represent the part of Y not explained by all the variables other than X. The X residuals represent the part of X not explained by other variables. The slope of the line fitted to the points in the added variable plot is equal to the regression coefficient when Y is regressed on all variables including X.
A strong linear relationship in the added variable plot indicates the increased importance of the contribution of X to the model already containing the other predictors.
Chatterjee, Samprit and Hadi, Ali. Regression Analysis by Example. 5th ed. N.p.: John Wiley & Sons, 2012. Print.
Kutner, MH, Nachtscheim CJ, Neter J and Li W., 2004, Applied Linear Statistical Models (5th edition). Chicago, IL., McGraw Hill/Irwin.
ols_plot_resid_regressor(), ols_plot_comp_plus_resid()
model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_plot_added_variable(model)model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_plot_added_variable(model)
The residual plus component plot indicates whether any non-linearity is present in the relationship between response and predictor variables and can suggest possible transformations for linearizing the data.
ols_plot_comp_plus_resid(model, print_plot = TRUE)ols_plot_comp_plus_resid(model, print_plot = TRUE)
model |
An object of class |
print_plot |
logical; if |
Chatterjee, Samprit and Hadi, Ali. Regression Analysis by Example. 5th ed. N.p.: John Wiley & Sons, 2012. Print.
Kutner, MH, Nachtscheim CJ, Neter J and Li W., 2004, Applied Linear Statistical Models (5th edition). Chicago, IL., McGraw Hill/Irwin.
ols_plot_added_variable(), ols_plot_resid_regressor()
model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_plot_comp_plus_resid(model)model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_plot_comp_plus_resid(model)
Bar Plot of cook's distance to detect observations that strongly influence fitted values of the model.
ols_plot_cooksd_bar(model, type = 1, threshold = NULL, print_plot = TRUE)ols_plot_cooksd_bar(model, type = 1, threshold = NULL, print_plot = TRUE)
model |
An object of class |
type |
An integer between 1 and 5 selecting one of the 5 methods for computing the threshold. |
threshold |
Threshold for detecting outliers. |
print_plot |
logical; if |
Cook's distance was introduced by American statistician R Dennis Cook in 1977. It is used to identify influential data points. It depends on both the residual and leverage i.e it takes it account both the x value and y value of the observation.
Steps to compute Cook's distance:
Delete observations one at a time.
Refit the regression model on remaining observations
examine how much all of the fitted values change when the ith observation is deleted.
A data point having a large cook's d indicates that the data point strongly influences the fitted values. There are several methods/formulas to compute the threshold used for detecting or classifying observations as outliers and we list them below.
Type 1 : 4 / n
Type 2 : 4 / (n - k - 1)
Type 3 : ~1
Type 4 : 1 / (n - k - 1)
Type 5 : 3 * mean(Vector of cook's distance values)
where n and k stand for
n: Number of observations
k: Number of predictors
ols_plot_cooksd_bar returns a list containing the
following components:
outliers |
a |
threshold |
|
model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_plot_cooksd_bar(model) ols_plot_cooksd_bar(model, type = 4) ols_plot_cooksd_bar(model, threshold = 0.2)model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_plot_cooksd_bar(model) ols_plot_cooksd_bar(model, type = 4) ols_plot_cooksd_bar(model, threshold = 0.2)
Chart of cook's distance to detect observations that strongly influence fitted values of the model.
ols_plot_cooksd_chart(model, type = 1, threshold = NULL, print_plot = TRUE)ols_plot_cooksd_chart(model, type = 1, threshold = NULL, print_plot = TRUE)
model |
An object of class |
type |
An integer between 1 and 5 selecting one of the 6 methods for computing the threshold. |
threshold |
Threshold for detecting outliers. |
print_plot |
logical; if |
Cook's distance was introduced by American statistician R Dennis Cook in 1977. It is used to identify influential data points. It depends on both the residual and leverage i.e it takes it account both the x value and y value of the observation.
Steps to compute Cook's distance:
Delete observations one at a time.
Refit the regression model on remaining observations
exmine how much all of the fitted values change when the ith observation is deleted.
A data point having a large cook's d indicates that the data point strongly influences the fitted values. There are several methods/formulas to compute the threshold used for detecting or classifying observations as outliers and we list them below.
Type 1 : 4 / n
Type 2 : 4 / (n - k - 1)
Type 3 : ~1
Type 4 : 1 / (n - k - 1)
Type 5 : 3 * mean(Vector of cook's distance values)
where n and k stand for
n: Number of observations
k: Number of predictors
ols_plot_cooksd_chart returns a list containing the
following components:
outliers |
a |
threshold |
|
model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_plot_cooksd_chart(model) ols_plot_cooksd_chart(model, type = 4) ols_plot_cooksd_chart(model, threshold = 0.2)model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_plot_cooksd_chart(model) ols_plot_cooksd_chart(model, type = 4) ols_plot_cooksd_chart(model, threshold = 0.2)
Panel of plots to detect influential observations using DFBETAs.
ols_plot_dfbetas(model, print_plot = TRUE)ols_plot_dfbetas(model, print_plot = TRUE)
model |
An object of class |
print_plot |
logical; if |
DFBETA measures the difference in each parameter estimate with and without
the influential point. There is a DFBETA for each data point i.e if there are
n observations and k variables, there will be DFBETAs. In
general, large values of DFBETAS indicate observations that are influential
in estimating a given parameter. Belsley, Kuh, and Welsch recommend 2 as a
general cutoff value to indicate influential observations and
as a size-adjusted cutoff.
list; ols_plot_dfbetas returns a list of data.frame (for intercept and each predictor)
with the observation number and DFBETA of observations that exceed the threshold for classifying
an observation as an outlier/influential observation.
Belsley, David A.; Kuh, Edwin; Welsh, Roy E. (1980). Regression Diagnostics: Identifying Influential Data and Sources of Collinearity.
Wiley Series in Probability and Mathematical Statistics. New York: John Wiley & Sons. pp. ISBN 0-471-05856-4.
model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_plot_dfbetas(model)model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_plot_dfbetas(model)
Plot for detecting influential observations using DFFITs.
ols_plot_dffits(model, size_adj_threshold = TRUE, print_plot = TRUE)ols_plot_dffits(model, size_adj_threshold = TRUE, print_plot = TRUE)
model |
An object of class |
size_adj_threshold |
logical; if |
print_plot |
logical; if |
DFFIT - difference in fits, is used to identify influential data points. It quantifies the number of standard deviations that the fitted value changes when the ith data point is omitted.
Steps to compute DFFITs:
Delete observations one at a time.
Refit the regression model on remaining observations
examine how much all of the fitted values change when the ith observation is deleted.
An observation is deemed influential if the absolute value of its DFFITS value is greater than:
A size-adjusted cutoff recommended by Belsley, Kuh, and Welsch is
and is used by default in olsrr.
where n is the number of observations and p is the number of predictors including intercept.
ols_plot_dffits returns a list containing the
following components:
outliers |
a |
threshold |
|
Belsley, David A.; Kuh, Edwin; Welsh, Roy E. (1980). Regression Diagnostics: Identifying Influential Data and Sources of Collinearity.
Wiley Series in Probability and Mathematical Statistics. New York: John Wiley & Sons. ISBN 0-471-05856-4.
model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_plot_dffits(model) ols_plot_dffits(model, size_adj_threshold = FALSE)model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_plot_dffits(model) ols_plot_dffits(model, size_adj_threshold = FALSE)
Panel of plots for regression diagnostics.
ols_plot_diagnostics(model, print_plot = TRUE)ols_plot_diagnostics(model, print_plot = TRUE)
model |
An object of class |
print_plot |
logical; if |
model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_plot_diagnostics(model)model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_plot_diagnostics(model)
Hadi's measure of influence based on the fact that influential observations can be present in either the response variable or in the predictors or both. The plot is used to detect influential observations based on Hadi's measure.
ols_plot_hadi(model, print_plot = TRUE)ols_plot_hadi(model, print_plot = TRUE)
model |
An object of class |
print_plot |
logical; if |
Chatterjee, Samprit and Hadi, Ali. Regression Analysis by Example. 5th ed. N.p.: John Wiley & Sons, 2012. Print.
model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_plot_hadi(model)model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_plot_hadi(model)
Plot of observed vs fitted values to assess the fit of the model.
ols_plot_obs_fit(model, print_plot = TRUE)ols_plot_obs_fit(model, print_plot = TRUE)
model |
An object of class |
print_plot |
logical; if |
Ideally, all your points should be close to a regressed diagonal line. Draw such a diagonal line within your graph and check out where the points lie. If your model had a high R Square, all the points would be close to this diagonal line. The lower the R Square, the weaker the Goodness of fit of your model, the more foggy or dispersed your points are from this diagonal line.
model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_plot_obs_fit(model)model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_plot_obs_fit(model)
Plot to demonstrate that the regression line always passes through mean of the response and predictor variables.
ols_plot_reg_line(response, predictor, print_plot = TRUE)ols_plot_reg_line(response, predictor, print_plot = TRUE)
response |
Response variable. |
predictor |
Predictor variable. |
print_plot |
logical; if |
ols_plot_reg_line(mtcars$mpg, mtcars$disp)ols_plot_reg_line(mtcars$mpg, mtcars$disp)
Box plot of residuals to examine if residuals are normally distributed.
ols_plot_resid_box(model, print_plot = TRUE)ols_plot_resid_box(model, print_plot = TRUE)
model |
An object of class |
print_plot |
logical; if |
Other residual diagnostics:
ols_plot_resid_fit(),
ols_plot_resid_hist(),
ols_plot_resid_qq(),
ols_test_correlation(),
ols_test_normality()
model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_plot_resid_box(model)model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_plot_resid_box(model)
Scatter plot of residuals on the y axis and fitted values on the x axis to detect non-linearity, unequal error variances, and outliers.
ols_plot_resid_fit(model, print_plot = TRUE)ols_plot_resid_fit(model, print_plot = TRUE)
model |
An object of class |
print_plot |
logical; if |
Characteristics of a well behaved residual vs fitted plot:
The residuals spread randomly around the 0 line indicating that the relationship is linear.
The residuals form an approximate horizontal band around the 0 line indicating homogeneity of error variance.
No one residual is visibly away from the random pattern of the residuals indicating that there are no outliers.
Other residual diagnostics:
ols_plot_resid_box(),
ols_plot_resid_hist(),
ols_plot_resid_qq(),
ols_test_correlation(),
ols_test_normality()
model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_plot_resid_fit(model)model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_plot_resid_fit(model)
Plot to detect non-linearity, influential observations and outliers.
ols_plot_resid_fit_spread(model, print_plot = TRUE) ols_plot_fm(model, print_plot = TRUE) ols_plot_resid_spread(model, print_plot = TRUE)ols_plot_resid_fit_spread(model, print_plot = TRUE) ols_plot_fm(model, print_plot = TRUE) ols_plot_resid_spread(model, print_plot = TRUE)
model |
An object of class |
print_plot |
logical; if |
Consists of side-by-side quantile plots of the centered fit and the residuals. It shows how much variation in the data is explained by the fit and how much remains in the residuals. For inappropriate models, the spread of the residuals in such a plot is often greater than the spread of the centered fit.
Cleveland, W. S. (1993). Visualizing Data. Summit, NJ: Hobart Press.
# model model <- lm(mpg ~ disp + hp + wt, data = mtcars) # residual fit spread plot ols_plot_resid_fit_spread(model) # fit mean plot ols_plot_fm(model) # residual spread plot ols_plot_resid_spread(model)# model model <- lm(mpg ~ disp + hp + wt, data = mtcars) # residual fit spread plot ols_plot_resid_fit_spread(model) # fit mean plot ols_plot_fm(model) # residual spread plot ols_plot_resid_spread(model)
Histogram of residuals for detecting violation of normality assumption.
ols_plot_resid_hist(model, print_plot = TRUE)ols_plot_resid_hist(model, print_plot = TRUE)
model |
An object of class |
print_plot |
logical; if |
Other residual diagnostics:
ols_plot_resid_box(),
ols_plot_resid_fit(),
ols_plot_resid_qq(),
ols_test_correlation(),
ols_test_normality()
model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_plot_resid_hist(model)model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_plot_resid_hist(model)
Graph for detecting outliers and/or observations with high leverage.
ols_plot_resid_lev(model, threshold = NULL, print_plot = TRUE)ols_plot_resid_lev(model, threshold = NULL, print_plot = TRUE)
model |
An object of class |
threshold |
Threshold for detecting outliers. Default is 2. |
print_plot |
logical; if |
ols_plot_resid_stud_fit(), ols_plot_resid_lev()
model <- lm(read ~ write + math + science, data = hsb) ols_plot_resid_lev(model) ols_plot_resid_lev(model, threshold = 3)model <- lm(read ~ write + math + science, data = hsb) ols_plot_resid_lev(model) ols_plot_resid_lev(model, threshold = 3)
Plot to aid in classifying unusual observations as high-leverage points, outliers, or a combination of both.
ols_plot_resid_pot(model, print_plot = TRUE)ols_plot_resid_pot(model, print_plot = TRUE)
model |
An object of class |
print_plot |
logical; if |
Chatterjee, Samprit and Hadi, Ali. Regression Analysis by Example. 5th ed. N.p.: John Wiley & Sons, 2012. Print.
model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_plot_resid_pot(model)model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_plot_resid_pot(model)
Graph for detecting violation of normality assumption.
ols_plot_resid_qq(model, print_plot = TRUE)ols_plot_resid_qq(model, print_plot = TRUE)
model |
An object of class |
print_plot |
logical; if |
Other residual diagnostics:
ols_plot_resid_box(),
ols_plot_resid_fit(),
ols_plot_resid_hist(),
ols_test_correlation(),
ols_test_normality()
model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_plot_resid_qq(model)model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_plot_resid_qq(model)
Graph to determine whether we should add a new predictor to the model already containing other predictors. The residuals from the model is regressed on the new predictor and if the plot shows non random pattern, you should consider adding the new predictor to the model.
ols_plot_resid_regressor(model, variable, print_plot = TRUE)ols_plot_resid_regressor(model, variable, print_plot = TRUE)
model |
An object of class |
variable |
New predictor to be added to the |
print_plot |
logical; if |
ols_plot_added_variable(), ols_plot_comp_plus_resid()
model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_plot_resid_regressor(model, 'drat')model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_plot_resid_regressor(model, 'drat')
Chart for identifying outliers.
ols_plot_resid_stand(model, threshold = NULL, print_plot = TRUE)ols_plot_resid_stand(model, threshold = NULL, print_plot = TRUE)
model |
An object of class |
threshold |
Threshold for detecting outliers. Default is 2. |
print_plot |
logical; if |
Standardized residual (internally studentized) is the residual divided by estimated standard deviation.
ols_plot_resid_stand returns a list containing the
following components:
outliers |
a |
for classifying an observation as an outlier
threshold |
|
model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_plot_resid_stand(model) ols_plot_resid_stand(model, threshold = 3)model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_plot_resid_stand(model) ols_plot_resid_stand(model, threshold = 3)
Graph for identifying outliers.
ols_plot_resid_stud(model, threshold = NULL, print_plot = TRUE)ols_plot_resid_stud(model, threshold = NULL, print_plot = TRUE)
model |
An object of class |
threshold |
Threshold for detecting outliers. Default is 3. |
print_plot |
logical; if |
Studentized deleted residuals (or externally studentized residuals) is the deleted residual divided by its estimated standard deviation. Studentized residuals are going to be more effective for detecting outlying Y observations than standardized residuals. If an observation has an externally studentized residual that is larger than 3 (in absolute value) we can call it an outlier.
ols_plot_resid_stud returns a list containing the
following components:
outliers |
a |
for classifying an observation as an outlier
threshold |
|
model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_plot_resid_stud(model) ols_plot_resid_stud(model, threshold = 2)model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_plot_resid_stud(model) ols_plot_resid_stud(model, threshold = 2)
Plot for detecting violation of assumptions about residuals such as non-linearity, constant variances and outliers. It can also be used to examine model fit.
ols_plot_resid_stud_fit(model, threshold = NULL, print_plot = TRUE)ols_plot_resid_stud_fit(model, threshold = NULL, print_plot = TRUE)
model |
An object of class |
threshold |
Threshold for detecting outliers. Default is 2. |
print_plot |
logical; if |
Studentized deleted residuals (or externally studentized residuals) is the deleted residual divided by its estimated standard deviation. Studentized residuals are going to be more effective for detecting outlying Y observations than standardized residuals. If an observation has an externally studentized residual that is larger than 2 (in absolute value) we can call it an outlier.
ols_plot_resid_stud_fit returns a list containing the
following components:
outliers |
a |
threshold |
|
ols_plot_resid_lev(), ols_plot_resid_stand(),
ols_plot_resid_stud()
model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_plot_resid_stud_fit(model) ols_plot_resid_stud_fit(model, threshold = 3)model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_plot_resid_stud_fit(model) ols_plot_resid_stud_fit(model, threshold = 3)
Panel of plots to explore and visualize the response variable.
ols_plot_response(model, print_plot = TRUE)ols_plot_response(model, print_plot = TRUE)
model |
An object of class |
print_plot |
logical; if |
model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_plot_response(model)model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_plot_response(model)
Use predicted rsquared to determine how well the model predicts responses for new observations. Larger values of predicted R2 indicate models of greater predictive ability.
ols_pred_rsq(model)ols_pred_rsq(model)
model |
An object of class |
Predicted rsquare of the model.
Other influence measures:
ols_hadi(),
ols_leverage(),
ols_press()
model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_pred_rsq(model)model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_pred_rsq(model)
Data for generating the added variable plots.
ols_prep_avplot_data(model)ols_prep_avplot_data(model)
model |
An object of class |
model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_prep_avplot_data(model)model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_prep_avplot_data(model)
Prepare data for cook's d bar plot.
ols_prep_cdplot_data(model, type = 1)ols_prep_cdplot_data(model, type = 1)
model |
An object of class |
type |
An integer between 1 and 5 selecting one of the 6 methods for computing the threshold. |
model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_prep_cdplot_data(model)model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_prep_cdplot_data(model)
Outlier data for cook's d bar plot.
ols_prep_cdplot_outliers(k)ols_prep_cdplot_outliers(k)
k |
Cooks' d bar plot data. |
model <- lm(mpg ~ disp + hp + wt, data = mtcars) k <- ols_prep_cdplot_data(model) ols_prep_cdplot_outliers(k)model <- lm(mpg ~ disp + hp + wt, data = mtcars) k <- ols_prep_cdplot_data(model) ols_prep_cdplot_outliers(k)
Prepares the data for dfbetas plot.
ols_prep_dfbeta_data(d, threshold)ols_prep_dfbeta_data(d, threshold)
d |
A |
threshold |
The threshold for outliers. |
model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) dfb <- dfbetas(model) n <- nrow(dfb) threshold <- 2 / sqrt(n) dbetas <- dfb[, 1] df_data <- data.frame(obs = seq_len(n), dbetas = dbetas) ols_prep_dfbeta_data(df_data, threshold)model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) dfb <- dfbetas(model) n <- nrow(dfb) threshold <- 2 / sqrt(n) dbetas <- dfb[, 1] df_data <- data.frame(obs = seq_len(n), dbetas = dbetas) ols_prep_dfbeta_data(df_data, threshold)
Data for identifying outliers in dfbetas plot.
ols_prep_dfbeta_outliers(d)ols_prep_dfbeta_outliers(d)
d |
A |
model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) dfb <- dfbetas(model) n <- nrow(dfb) threshold <- 2 / sqrt(n) dbetas <- dfb[, 1] df_data <- data.frame(obs = seq_len(n), dbetas = dbetas) d <- ols_prep_dfbeta_data(df_data, threshold) ols_prep_dfbeta_outliers(d)model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) dfb <- dfbetas(model) n <- nrow(dfb) threshold <- 2 / sqrt(n) dbetas <- dfb[, 1] df_data <- data.frame(obs = seq_len(n), dbetas = dbetas) d <- ols_prep_dfbeta_data(df_data, threshold) ols_prep_dfbeta_outliers(d)
Generates data for deleted studentized residual vs fitted plot.
ols_prep_dsrvf_data(model, threshold = NULL)ols_prep_dsrvf_data(model, threshold = NULL)
model |
An object of class |
threshold |
Threshold for detecting outliers. Default is 2. |
model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_prep_dsrvf_data(model) ols_prep_dsrvf_data(model, threshold = 3)model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_prep_dsrvf_data(model) ols_prep_dsrvf_data(model, threshold = 3)
Identify outliers in cook's d plot.
ols_prep_outlier_obs(k)ols_prep_outlier_obs(k)
k |
Cooks' d bar plot data. |
model <- lm(mpg ~ disp + hp + wt, data = mtcars) k <- ols_prep_cdplot_data(model) ols_prep_outlier_obs(k)model <- lm(mpg ~ disp + hp + wt, data = mtcars) k <- ols_prep_cdplot_data(model) ols_prep_outlier_obs(k)
Regress a predictor in the model on all the other predictors.
ols_prep_regress_x(data, i)ols_prep_regress_x(data, i)
data |
A |
i |
A numeric vector (indicates the predictor in the model). |
model <- lm(mpg ~ disp + hp + wt, data = mtcars) data <- ols_prep_avplot_data(model) ols_prep_regress_x(data, 1)model <- lm(mpg ~ disp + hp + wt, data = mtcars) data <- ols_prep_avplot_data(model) ols_prep_regress_x(data, 1)
Regress y on all the predictors except the ith predictor.
ols_prep_regress_y(data, i)ols_prep_regress_y(data, i)
data |
A |
i |
A numeric vector (indicates the predictor in the model). |
model <- lm(mpg ~ disp + hp + wt, data = mtcars) data <- ols_prep_avplot_data(model) ols_prep_regress_y(data, 1)model <- lm(mpg ~ disp + hp + wt, data = mtcars) data <- ols_prep_avplot_data(model) ols_prep_regress_y(data, 1)
Data for generating residual fit spread plot.
ols_prep_rfsplot_fmdata(model) ols_prep_rfsplot_rsdata(model)ols_prep_rfsplot_fmdata(model) ols_prep_rfsplot_rsdata(model)
model |
An object of class |
model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_prep_rfsplot_fmdata(model) ols_prep_rfsplot_rsdata(model)model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_prep_rfsplot_fmdata(model) ols_prep_rfsplot_rsdata(model)
Generates data for studentized resiudual vs leverage plot.
ols_prep_rstudlev_data(model, threshold = NULL)ols_prep_rstudlev_data(model, threshold = NULL)
model |
An object of class |
threshold |
Threshold for detecting outliers. Default is 2. |
model <- lm(read ~ write + math + science, data = hsb) ols_prep_rstudlev_data(model) ols_prep_rstudlev_data(model, threshold = 3)model <- lm(read ~ write + math + science, data = hsb) ols_prep_rstudlev_data(model) ols_prep_rstudlev_data(model, threshold = 3)
Data for generating residual vs regressor plot.
ols_prep_rvsrplot_data(model)ols_prep_rvsrplot_data(model)
model |
An object of class |
model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_prep_rvsrplot_data(model)model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_prep_rvsrplot_data(model)
Generates data for standardized residual chart.
ols_prep_srchart_data(model, threshold = NULL)ols_prep_srchart_data(model, threshold = NULL)
model |
An object of class |
threshold |
Threshold for detecting outliers. Default is 2. |
model <- lm(read ~ write + math + science, data = hsb) ols_prep_srchart_data(model) ols_prep_srchart_data(model, threshold = 3)model <- lm(read ~ write + math + science, data = hsb) ols_prep_srchart_data(model) ols_prep_srchart_data(model, threshold = 3)
Generates data for studentized residual plot.
ols_prep_srplot_data(model, threshold = NULL)ols_prep_srplot_data(model, threshold = NULL)
model |
An object of class |
threshold |
Threshold for detecting outliers. Default is 3. |
model <- lm(read ~ write + math + science, data = hsb) ols_prep_srplot_data(model)model <- lm(read ~ write + math + science, data = hsb) ols_prep_srplot_data(model)
PRESS (prediction sum of squares) tells you how well the model will predict new data.
ols_press(model)ols_press(model)
model |
An object of class |
The prediction sum of squares (PRESS) is the sum of squares of the prediction error. Each fitted to obtain the predicted value for the ith observation. Use PRESS to assess your model's predictive ability. Usually, the smaller the PRESS value, the better the model's predictive ability.
Predicted sum of squares of the model.
Kutner, MH, Nachtscheim CJ, Neter J and Li W., 2004, Applied Linear Statistical Models (5th edition). Chicago, IL., McGraw Hill/Irwin.
Other influence measures:
ols_hadi(),
ols_leverage(),
ols_pred_rsq()
model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_press(model)model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_press(model)
Assess how much of the error in prediction is due to lack of model fit.
ols_pure_error_anova(model, ...)ols_pure_error_anova(model, ...)
model |
An object of class |
... |
Other parameters. |
The residual sum of squares resulting from a regression can be decomposed into 2 components:
Due to lack of fit
Due to random variation
If most of the error is due to lack of fit and not just random error, the model should be discarded and a new model must be built.
ols_pure_error_anova returns an object of class
"ols_pure_error_anova". An object of class "ols_pure_error_anova" is a
list containing the following components:
lackoffit |
lack of fit sum of squares |
pure_error |
pure error sum of squares |
rss |
regression sum of squares |
ess |
error sum of squares |
total |
total sum of squares |
rms |
regression mean square |
ems |
error mean square |
lms |
lack of fit mean square |
pms |
pure error mean square |
rf |
f statistic |
lf |
lack of fit f statistic |
pr |
p-value of f statistic |
pl |
p-value pf lack of fit f statistic |
mpred |
|
df_rss |
regression sum of squares degrees of freedom |
df_ess |
error sum of squares degrees of freedom |
df_lof |
lack of fit degrees of freedom |
df_error |
pure error degrees of freedom |
final |
data.frame; contains computed values used for the lack of fit f test |
resp |
character vector; name of |
preds |
character vector; name of |
The lack of fit F test works only with simple linear regression. Moreover, it is important that the data contains repeat observations i.e. replicates for at least one of the values of the predictor x. This test generally only applies to datasets with plenty of replicates.
Kutner, MH, Nachtscheim CJ, Neter J and Li W., 2004, Applied Linear Statistical Models (5th edition). Chicago, IL., McGraw Hill/Irwin.
model <- lm(mpg ~ disp, data = mtcars) ols_pure_error_anova(model)model <- lm(mpg ~ disp, data = mtcars) ols_pure_error_anova(model)
Ordinary least squares regression.
ols_regress(object, ...) ## S3 method for class 'lm' ols_regress(object, ...)ols_regress(object, ...) ## S3 method for class 'lm' ols_regress(object, ...)
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
|
... |
Other inputs. |
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 statistis |
p |
p-value for |
n |
number of predictors including intercept |
betas |
betas; estimated coefficients |
sbetas |
standardized betas |
std_errors |
standard errors |
tvalues |
t values |
pvalues |
p-value of |
df |
degrees of freedom of |
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 |
If the model includes interaction terms, the standardized betas are computed after scaling and centering the predictors.
https://www.ssc.wisc.edu/~hemken/Stataworkshops/stdBeta/Getting%20Standardized%20Coefficients%20Right.pdf
ols_regress(mpg ~ disp + hp + wt, data = mtcars) # if model includes interaction terms set iterm to TRUE ols_regress(mpg ~ disp * wt, data = mtcars, iterm = TRUE)ols_regress(mpg ~ disp + hp + wt, data = mtcars) # if model includes interaction terms set iterm to TRUE ols_regress(mpg ~ disp * wt, data = mtcars, iterm = TRUE)
Bayesian information criterion for model selection.
ols_sbc(model, method = c("R", "STATA", "SAS"))ols_sbc(model, method = c("R", "STATA", "SAS"))
model |
An object of class |
method |
A character vector; specify the method to compute BIC. Valid options include R, STATA and SAS. |
SBC provides a means for model selection. Given a collection of models for the data, SBC estimates the quality of each model, relative to each of the other models. R and STATA use loglikelihood to compute SBC. SAS uses residual sum of squares. Below is the formula in each case:
R & STATA
SAS
where n is the sample size and p is the number of model parameters including intercept.
The bayesian information criterion of the model.
Schwarz, G. (1978). “Estimating the Dimension of a Model.” Annals of Statistics 6:461–464.
Judge, G. G., Griffiths, W. E., Hill, R. C., and Lee, T.-C. (1980). The Theory and Practice of Econometrics. New York: John Wiley & Sons.
Other model selection criteria:
ols_aic(),
ols_apc(),
ols_fpe(),
ols_hsp(),
ols_mallows_cp(),
ols_msep(),
ols_sbic()
# using R computation method model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_sbc(model) # using STATA computation method model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_sbc(model, method = 'STATA') # using SAS computation method model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_sbc(model, method = 'SAS')# using R computation method model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_sbc(model) # using STATA computation method model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_sbc(model, method = 'STATA') # using SAS computation method model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_sbc(model, method = 'SAS')
Sawa's bayesian information criterion for model selection.
ols_sbic(model, full_model)ols_sbic(model, full_model)
model |
An object of class |
full_model |
An object of class |
Sawa (1978) developed a model selection criterion that was derived from a Bayesian modification of the AIC criterion. Sawa's Bayesian Information Criterion (BIC) is a function of the number of observations n, the SSE, the pure error variance fitting the full model, and the number of independent variables including the intercept.
where , n is the sample size, p is the number of model parameters including intercept
SSE is the residual sum of squares.
Sawa's Bayesian Information Criterion
Sawa, T. (1978). “Information Criteria for Discriminating among Alternative Regression Models.” Econometrica 46:1273–1282.
Judge, G. G., Griffiths, W. E., Hill, R. C., and Lee, T.-C. (1980). The Theory and Practice of Econometrics. New York: John Wiley & Sons.
Other model selection criteria:
ols_aic(),
ols_apc(),
ols_fpe(),
ols_hsp(),
ols_mallows_cp(),
ols_msep(),
ols_sbc()
full_model <- lm(mpg ~ ., data = mtcars) model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_sbic(model, full_model)full_model <- lm(mpg ~ ., data = mtcars) model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_sbic(model, full_model)
Fits all regressions involving one regressor, two regressors, three regressors, and so on. It tests all possible subsets of the set of potential independent variables.
ols_step_all_possible(model, ...) ## Default S3 method: ols_step_all_possible(model, max_order = NULL, ...) ## S3 method for class 'ols_step_all_possible' plot(x, model = NA, print_plot = TRUE, ...)ols_step_all_possible(model, ...) ## Default S3 method: ols_step_all_possible(model, max_order = NULL, ...) ## S3 method for class 'ols_step_all_possible' plot(x, model = NA, print_plot = TRUE, ...)
model |
An object of class |
... |
Other arguments. |
max_order |
Maximum subset order. |
x |
An object of class |
print_plot |
logical; if |
ols_step_all_possible returns an object of class "ols_step_all_possible".
An object of class "ols_step_all_possible" is a data frame containing the
following components:
mindex |
model index |
n |
number of predictors |
predictors |
predictors in the model |
rsquare |
rsquare of the model |
adjr |
adjusted rsquare of the model |
rmse |
root mean squared error of the model |
predrsq |
predicted rsquare of the model |
cp |
mallow's Cp |
aic |
akaike information criteria |
sbic |
sawa bayesian information criteria |
sbc |
schwarz bayes information criteria |
msep |
estimated MSE of prediction, assuming multivariate normality |
fpe |
final prediction error |
apc |
amemiya prediction criteria |
hsp |
hocking's Sp |
Mendenhall William and Sinsich Terry, 2012, A Second Course in Statistics Regression Analysis (7th edition). Prentice Hall
model <- lm(mpg ~ disp + hp, data = mtcars) k <- ols_step_all_possible(model) k # plot plot(k) # maximum subset model <- lm(mpg ~ disp + hp + drat + wt + qsec, data = mtcars) ols_step_all_possible(model, max_order = 3)model <- lm(mpg ~ disp + hp, data = mtcars) k <- ols_step_all_possible(model) k # plot plot(k) # maximum subset model <- lm(mpg ~ disp + hp + drat + wt + qsec, data = mtcars) ols_step_all_possible(model, max_order = 3)
Returns the coefficients for each variable from each model.
ols_step_all_possible_betas(object, ...)ols_step_all_possible_betas(object, ...)
object |
An object of class |
... |
Other arguments. |
ols_step_all_possible_betas returns a data.frame containing:
model_index |
model number |
predictor |
predictor |
beta_coef |
coefficient for the predictor |
## Not run: model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_step_all_possible_betas(model) ## End(Not run)## Not run: model <- lm(mpg ~ disp + hp + wt, data = mtcars) ols_step_all_possible_betas(model) ## End(Not run)
Build regression model from a set of candidate predictor variables by removing predictors based on adjusted r-squared, in a stepwise manner until there is no variable left to remove any more.
ols_step_backward_adj_r2(model, ...) ## Default S3 method: ols_step_backward_adj_r2( model, include = NULL, exclude = NULL, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_backward_adj_r2' plot(x, print_plot = TRUE, details = TRUE, digits = 3, ...)ols_step_backward_adj_r2(model, ...) ## Default S3 method: ols_step_backward_adj_r2( model, include = NULL, exclude = NULL, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_backward_adj_r2' plot(x, print_plot = TRUE, details = TRUE, digits = 3, ...)
model |
An object of class |
... |
Other arguments. |
include |
Character or numeric vector; variables to be included in selection process. |
exclude |
Character or numeric vector; variables to be excluded from selection process. |
progress |
Logical; if |
details |
Logical; if |
x |
An object of class |
print_plot |
logical; if |
digits |
Number of decimal places to display. |
List containing the following components:
model |
final model; an object of class |
metrics |
selection metrics |
others |
list; info used for plotting and printing |
Venables, W. N. and Ripley, B. D. (2002) Modern Applied Statistics with S. Fourth edition. Springer.
Other backward selection procedures:
ols_step_backward_aic(),
ols_step_backward_p(),
ols_step_backward_r2(),
ols_step_backward_sbc(),
ols_step_backward_sbic()
# stepwise backward regression model <- lm(y ~ ., data = surgical) ols_step_backward_adj_r2(model) # final model and selection metrics k <- ols_step_backward_aic(model) k$metrics k$model # include or exclude variable # force variables to be included in the selection process ols_step_backward_adj_r2(model, include = c("alc_mod", "gender")) # use index of variable instead of name ols_step_backward_adj_r2(model, include = c(7, 6)) # force variable to be excluded from selection process ols_step_backward_adj_r2(model, exclude = c("alc_heavy", "bcs")) # use index of variable instead of name ols_step_backward_adj_r2(model, exclude = c(8, 1))# stepwise backward regression model <- lm(y ~ ., data = surgical) ols_step_backward_adj_r2(model) # final model and selection metrics k <- ols_step_backward_aic(model) k$metrics k$model # include or exclude variable # force variables to be included in the selection process ols_step_backward_adj_r2(model, include = c("alc_mod", "gender")) # use index of variable instead of name ols_step_backward_adj_r2(model, include = c(7, 6)) # force variable to be excluded from selection process ols_step_backward_adj_r2(model, exclude = c("alc_heavy", "bcs")) # use index of variable instead of name ols_step_backward_adj_r2(model, exclude = c(8, 1))
Build regression model from a set of candidate predictor variables by removing predictors based on akaike information criterion, in a stepwise manner until there is no variable left to remove any more.
ols_step_backward_aic(model, ...) ## Default S3 method: ols_step_backward_aic( model, include = NULL, exclude = NULL, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_backward_aic' plot(x, print_plot = TRUE, details = TRUE, digits = 3, ...)ols_step_backward_aic(model, ...) ## Default S3 method: ols_step_backward_aic( model, include = NULL, exclude = NULL, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_backward_aic' plot(x, print_plot = TRUE, details = TRUE, digits = 3, ...)
model |
An object of class |
... |
Other arguments. |
include |
Character or numeric vector; variables to be included in selection process. |
exclude |
Character or numeric vector; variables to be excluded from selection process. |
progress |
Logical; if |
details |
Logical; if |
x |
An object of class |
print_plot |
logical; if |
digits |
Number of decimal places to display. |
List containing the following components:
model |
final model; an object of class |
metrics |
selection metrics |
others |
list; info used for plotting and printing |
Venables, W. N. and Ripley, B. D. (2002) Modern Applied Statistics with S. Fourth edition. Springer.
Other backward selection procedures:
ols_step_backward_adj_r2(),
ols_step_backward_p(),
ols_step_backward_r2(),
ols_step_backward_sbc(),
ols_step_backward_sbic()
# stepwise backward regression model <- lm(y ~ ., data = surgical) ols_step_backward_aic(model) # stepwise backward regression plot model <- lm(y ~ ., data = surgical) k <- ols_step_backward_aic(model) plot(k) # selection metrics k$metrics # final model k$model # include or exclude variable # force variables to be included in the selection process ols_step_backward_aic(model, include = c("alc_mod", "gender")) # use index of variable instead of name ols_step_backward_aic(model, include = c(7, 6)) # force variable to be excluded from selection process ols_step_backward_aic(model, exclude = c("alc_heavy", "bcs")) # use index of variable instead of name ols_step_backward_aic(model, exclude = c(8, 1))# stepwise backward regression model <- lm(y ~ ., data = surgical) ols_step_backward_aic(model) # stepwise backward regression plot model <- lm(y ~ ., data = surgical) k <- ols_step_backward_aic(model) plot(k) # selection metrics k$metrics # final model k$model # include or exclude variable # force variables to be included in the selection process ols_step_backward_aic(model, include = c("alc_mod", "gender")) # use index of variable instead of name ols_step_backward_aic(model, include = c(7, 6)) # force variable to be excluded from selection process ols_step_backward_aic(model, exclude = c("alc_heavy", "bcs")) # use index of variable instead of name ols_step_backward_aic(model, exclude = c(8, 1))
Build regression model from a set of candidate predictor variables by removing predictors based on p values, in a stepwise manner until there is no variable left to remove any more.
ols_step_backward_p(model, ...) ## Default S3 method: ols_step_backward_p( model, p_val = 0.3, include = NULL, exclude = NULL, hierarchical = FALSE, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_backward_p' plot(x, model = NA, print_plot = TRUE, details = TRUE, ...)ols_step_backward_p(model, ...) ## Default S3 method: ols_step_backward_p( model, p_val = 0.3, include = NULL, exclude = NULL, hierarchical = FALSE, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_backward_p' plot(x, model = NA, print_plot = TRUE, details = TRUE, ...)
model |
An object of class |
... |
Other inputs. |
p_val |
p value; variables with p more than |
include |
Character or numeric vector; variables to be included in selection process. |
exclude |
Character or numeric vector; variables to be excluded from selection process. |
hierarchical |
Logical; if |
progress |
Logical; if |
details |
Logical; if |
x |
An object of class |
print_plot |
logical; if |
ols_step_backward_p returns an object of class "ols_step_backward_p".
An object of class "ols_step_backward_p" is a list containing the
following components:
model |
final model; an object of class |
metrics |
selection metrics |
Chatterjee, Samprit and Hadi, Ali. Regression Analysis by Example. 5th ed. N.p.: John Wiley & Sons, 2012. Print.
Other backward selection procedures:
ols_step_backward_adj_r2(),
ols_step_backward_aic(),
ols_step_backward_r2(),
ols_step_backward_sbc(),
ols_step_backward_sbic()
# stepwise backward regression model <- lm(y ~ ., data = surgical) ols_step_backward_p(model) # stepwise backward regression plot model <- lm(y ~ ., data = surgical) k <- ols_step_backward_p(model) plot(k) # selection metrics k$metrics # final model k$model # include or exclude variables # force variable to be included in selection process ols_step_backward_p(model, include = c("age", "alc_mod")) # use index of variable instead of name ols_step_backward_p(model, include = c(5, 7)) # force variable to be excluded from selection process ols_step_backward_p(model, exclude = c("pindex")) # use index of variable instead of name ols_step_backward_p(model, exclude = c(2)) # hierarchical selection model <- lm(y ~ bcs + alc_heavy + pindex + age + alc_mod, data = surgical) ols_step_backward_p(model, 0.1, hierarchical = TRUE) # plot k <- ols_step_backward_p(model, 0.1, hierarchical = TRUE) plot(k)# stepwise backward regression model <- lm(y ~ ., data = surgical) ols_step_backward_p(model) # stepwise backward regression plot model <- lm(y ~ ., data = surgical) k <- ols_step_backward_p(model) plot(k) # selection metrics k$metrics # final model k$model # include or exclude variables # force variable to be included in selection process ols_step_backward_p(model, include = c("age", "alc_mod")) # use index of variable instead of name ols_step_backward_p(model, include = c(5, 7)) # force variable to be excluded from selection process ols_step_backward_p(model, exclude = c("pindex")) # use index of variable instead of name ols_step_backward_p(model, exclude = c(2)) # hierarchical selection model <- lm(y ~ bcs + alc_heavy + pindex + age + alc_mod, data = surgical) ols_step_backward_p(model, 0.1, hierarchical = TRUE) # plot k <- ols_step_backward_p(model, 0.1, hierarchical = TRUE) plot(k)
Build regression model from a set of candidate predictor variables by removing predictors based on r-squared, in a stepwise manner until there is no variable left to remove any more.
ols_step_backward_r2(model, ...) ## Default S3 method: ols_step_backward_r2( model, include = NULL, exclude = NULL, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_backward_r2' plot(x, print_plot = TRUE, details = TRUE, digits = 3, ...)ols_step_backward_r2(model, ...) ## Default S3 method: ols_step_backward_r2( model, include = NULL, exclude = NULL, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_backward_r2' plot(x, print_plot = TRUE, details = TRUE, digits = 3, ...)
model |
An object of class |
... |
Other arguments. |
include |
Character or numeric vector; variables to be included in selection process. |
exclude |
Character or numeric vector; variables to be excluded from selection process. |
progress |
Logical; if |
details |
Logical; if |
x |
An object of class |
print_plot |
logical; if |
digits |
Number of decimal places to display. |
List containing the following components:
model |
final model; an object of class |
metrics |
selection metrics |
others |
list; info used for plotting and printing |
Venables, W. N. and Ripley, B. D. (2002) Modern Applied Statistics with S. Fourth edition. Springer.
Other backward selection procedures:
ols_step_backward_adj_r2(),
ols_step_backward_aic(),
ols_step_backward_p(),
ols_step_backward_sbc(),
ols_step_backward_sbic()
# stepwise backward regression model <- lm(y ~ ., data = surgical) ols_step_backward_r2(model) # final model and selection metrics k <- ols_step_backward_aic(model) k$metrics k$model # include or exclude variable # force variables to be included in the selection process ols_step_backward_r2(model, include = c("alc_mod", "gender")) # use index of variable instead of name ols_step_backward_r2(model, include = c(7, 6)) # force variable to be excluded from selection process ols_step_backward_r2(model, exclude = c("alc_heavy", "bcs")) # use index of variable instead of name ols_step_backward_r2(model, exclude = c(8, 1))# stepwise backward regression model <- lm(y ~ ., data = surgical) ols_step_backward_r2(model) # final model and selection metrics k <- ols_step_backward_aic(model) k$metrics k$model # include or exclude variable # force variables to be included in the selection process ols_step_backward_r2(model, include = c("alc_mod", "gender")) # use index of variable instead of name ols_step_backward_r2(model, include = c(7, 6)) # force variable to be excluded from selection process ols_step_backward_r2(model, exclude = c("alc_heavy", "bcs")) # use index of variable instead of name ols_step_backward_r2(model, exclude = c(8, 1))
Build regression model from a set of candidate predictor variables by removing predictors based on schwarz bayesian criterion, in a stepwise manner until there is no variable left to remove any more.
ols_step_backward_sbc(model, ...) ## Default S3 method: ols_step_backward_sbc( model, include = NULL, exclude = NULL, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_backward_sbc' plot(x, print_plot = TRUE, details = TRUE, digits = 3, ...)ols_step_backward_sbc(model, ...) ## Default S3 method: ols_step_backward_sbc( model, include = NULL, exclude = NULL, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_backward_sbc' plot(x, print_plot = TRUE, details = TRUE, digits = 3, ...)
model |
An object of class |
... |
Other arguments. |
include |
Character or numeric vector; variables to be included in selection process. |
exclude |
Character or numeric vector; variables to be excluded from selection process. |
progress |
Logical; if |
details |
Logical; if |
x |
An object of class |
print_plot |
logical; if |
digits |
Number of decimal places to display. |
List containing the following components:
model |
final model; an object of class |
metrics |
selection metrics |
others |
list; info used for plotting and printing |
Venables, W. N. and Ripley, B. D. (2002) Modern Applied Statistics with S. Fourth edition. Springer.
Other backward selection procedures:
ols_step_backward_adj_r2(),
ols_step_backward_aic(),
ols_step_backward_p(),
ols_step_backward_r2(),
ols_step_backward_sbic()
# stepwise backward regression model <- lm(y ~ ., data = surgical) ols_step_backward_sbc(model) # stepwise backward regression plot model <- lm(y ~ ., data = surgical) k <- ols_step_backward_sbc(model) plot(k) # selection metrics k$metrics # final model k$model # include or exclude variable # force variables to be included in the selection process ols_step_backward_sbc(model, include = c("alc_mod", "gender")) # use index of variable instead of name ols_step_backward_sbc(model, include = c(7, 6)) # force variable to be excluded from selection process ols_step_backward_sbc(model, exclude = c("alc_heavy", "bcs")) # use index of variable instead of name ols_step_backward_sbc(model, exclude = c(8, 1))# stepwise backward regression model <- lm(y ~ ., data = surgical) ols_step_backward_sbc(model) # stepwise backward regression plot model <- lm(y ~ ., data = surgical) k <- ols_step_backward_sbc(model) plot(k) # selection metrics k$metrics # final model k$model # include or exclude variable # force variables to be included in the selection process ols_step_backward_sbc(model, include = c("alc_mod", "gender")) # use index of variable instead of name ols_step_backward_sbc(model, include = c(7, 6)) # force variable to be excluded from selection process ols_step_backward_sbc(model, exclude = c("alc_heavy", "bcs")) # use index of variable instead of name ols_step_backward_sbc(model, exclude = c(8, 1))
Build regression model from a set of candidate predictor variables by removing predictors based on sawa bayesian criterion, in a stepwise manner until there is no variable left to remove any more.
ols_step_backward_sbic(model, ...) ## Default S3 method: ols_step_backward_sbic( model, include = NULL, exclude = NULL, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_backward_sbic' plot(x, print_plot = TRUE, details = TRUE, digits = 3, ...)ols_step_backward_sbic(model, ...) ## Default S3 method: ols_step_backward_sbic( model, include = NULL, exclude = NULL, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_backward_sbic' plot(x, print_plot = TRUE, details = TRUE, digits = 3, ...)
model |
An object of class |
... |
Other arguments. |
include |
Character or numeric vector; variables to be included in selection process. |
exclude |
Character or numeric vector; variables to be excluded from selection process. |
progress |
Logical; if |
details |
Logical; if |
x |
An object of class |
print_plot |
logical; if |
digits |
Number of decimal places to display. |
List containing the following components:
model |
final model; an object of class |
metrics |
selection metrics |
others |
list; info used for plotting and printing |
Venables, W. N. and Ripley, B. D. (2002) Modern Applied Statistics with S. Fourth edition. Springer.
Other backward selection procedures:
ols_step_backward_adj_r2(),
ols_step_backward_aic(),
ols_step_backward_p(),
ols_step_backward_r2(),
ols_step_backward_sbc()
# stepwise backward regression model <- lm(y ~ ., data = surgical) ols_step_backward_sbic(model) # stepwise backward regression plot model <- lm(y ~ ., data = surgical) k <- ols_step_backward_sbic(model) plot(k) # selection metrics k$metrics # final model k$model # include or exclude variable # force variables to be included in the selection process ols_step_backward_sbic(model, include = c("alc_mod", "gender")) # use index of variable instead of name ols_step_backward_sbic(model, include = c(7, 6)) # force variable to be excluded from selection process ols_step_backward_sbic(model, exclude = c("alc_heavy", "bcs")) # use index of variable instead of name ols_step_backward_sbic(model, exclude = c(8, 1))# stepwise backward regression model <- lm(y ~ ., data = surgical) ols_step_backward_sbic(model) # stepwise backward regression plot model <- lm(y ~ ., data = surgical) k <- ols_step_backward_sbic(model) plot(k) # selection metrics k$metrics # final model k$model # include or exclude variable # force variables to be included in the selection process ols_step_backward_sbic(model, include = c("alc_mod", "gender")) # use index of variable instead of name ols_step_backward_sbic(model, include = c(7, 6)) # force variable to be excluded from selection process ols_step_backward_sbic(model, exclude = c("alc_heavy", "bcs")) # use index of variable instead of name ols_step_backward_sbic(model, exclude = c(8, 1))
Select the subset of predictors that do the best at meeting some well-defined objective criterion, such as having the largest R2 value or the smallest MSE, Mallow's Cp or AIC. The default metric used for selecting the model is R2 but the user can choose any of the other available metrics.
ols_step_best_subset(model, ...) ## Default S3 method: ols_step_best_subset( model, max_order = NULL, include = NULL, exclude = NULL, metric = c("rsquare", "adjr", "predrsq", "cp", "aic", "sbic", "sbc", "msep", "fpe", "apc", "hsp"), ... ) ## S3 method for class 'ols_step_best_subset' plot(x, model = NA, print_plot = TRUE, ...)ols_step_best_subset(model, ...) ## Default S3 method: ols_step_best_subset( model, max_order = NULL, include = NULL, exclude = NULL, metric = c("rsquare", "adjr", "predrsq", "cp", "aic", "sbic", "sbc", "msep", "fpe", "apc", "hsp"), ... ) ## S3 method for class 'ols_step_best_subset' plot(x, model = NA, print_plot = TRUE, ...)
model |
An object of class |
... |
Other inputs. |
max_order |
Maximum subset order. |
include |
Character or numeric vector; variables to be included in selection process. |
exclude |
Character or numeric vector; variables to be excluded from selection process. |
metric |
Metric to select model. |
x |
An object of class |
print_plot |
logical; if |
ols_step_best_subset returns an object of class "ols_step_best_subset".
An object of class "ols_step_best_subset" is a list containing the following:
metrics |
selection metrics |
Kutner, MH, Nachtscheim CJ, Neter J and Li W., 2004, Applied Linear Statistical Models (5th edition). Chicago, IL., McGraw Hill/Irwin.
model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_step_best_subset(model) ols_step_best_subset(model, metric = "adjr") ols_step_best_subset(model, metric = "cp") # maximum subset model <- lm(mpg ~ disp + hp + drat + wt + qsec, data = mtcars) ols_step_best_subset(model, max_order = 3) # plot model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) k <- ols_step_best_subset(model) plot(k) # return only models including `qsec` ols_step_best_subset(model, include = c("qsec")) # exclude `hp` from selection process ols_step_best_subset(model, exclude = c("hp"))model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_step_best_subset(model) ols_step_best_subset(model, metric = "adjr") ols_step_best_subset(model, metric = "cp") # maximum subset model <- lm(mpg ~ disp + hp + drat + wt + qsec, data = mtcars) ols_step_best_subset(model, max_order = 3) # plot model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) k <- ols_step_best_subset(model) plot(k) # return only models including `qsec` ols_step_best_subset(model, include = c("qsec")) # exclude `hp` from selection process ols_step_best_subset(model, exclude = c("hp"))
Build regression model from a set of candidate predictor variables by entering and removing predictors based on adjusted r-squared, in a stepwise manner until there is no variable left to enter or remove any more.
ols_step_both_adj_r2(model, ...) ## Default S3 method: ols_step_both_adj_r2( model, include = NULL, exclude = NULL, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_both_adj_r2' plot(x, print_plot = TRUE, details = TRUE, digits = 3, ...)ols_step_both_adj_r2(model, ...) ## Default S3 method: ols_step_both_adj_r2( model, include = NULL, exclude = NULL, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_both_adj_r2' plot(x, print_plot = TRUE, details = TRUE, digits = 3, ...)
model |
An object of class |
... |
Other arguments. |
include |
Character or numeric vector; variables to be included in selection process. |
exclude |
Character or numeric vector; variables to be excluded from selection process. |
progress |
Logical; if |
details |
Logical; if |
x |
An object of class |
print_plot |
logical; if |
digits |
Number of decimal places to display. |
List containing the following components:
model |
final model; an object of class |
metrics |
selection metrics |
others |
list; info used for plotting and printing |
Venables, W. N. and Ripley, B. D. (2002) Modern Applied Statistics with S. Fourth edition. Springer.
Other both direction selection procedures:
ols_step_both_aic(),
ols_step_both_r2(),
ols_step_both_sbc(),
ols_step_both_sbic()
## Not run: # stepwise regression model <- lm(y ~ ., data = stepdata) ols_step_both_adj_r2(model) # stepwise regression plot model <- lm(y ~ ., data = stepdata) k <- ols_step_both_adj_r2(model) plot(k) # selection metrics k$metrics # final model k$model # include or exclude variables # force variable to be included in selection process model <- lm(y ~ ., data = stepdata) ols_step_both_adj_r2(model, include = c("x6")) # use index of variable instead of name ols_step_both_adj_r2(model, include = c(6)) # force variable to be excluded from selection process ols_step_both_adj_r2(model, exclude = c("x2")) # use index of variable instead of name ols_step_both_adj_r2(model, exclude = c(2)) # include & exclude variables in the selection process ols_step_both_adj_r2(model, include = c("x6"), exclude = c("x2")) # use index of variable instead of name ols_step_both_adj_r2(model, include = c(6), exclude = c(2)) ## End(Not run)## Not run: # stepwise regression model <- lm(y ~ ., data = stepdata) ols_step_both_adj_r2(model) # stepwise regression plot model <- lm(y ~ ., data = stepdata) k <- ols_step_both_adj_r2(model) plot(k) # selection metrics k$metrics # final model k$model # include or exclude variables # force variable to be included in selection process model <- lm(y ~ ., data = stepdata) ols_step_both_adj_r2(model, include = c("x6")) # use index of variable instead of name ols_step_both_adj_r2(model, include = c(6)) # force variable to be excluded from selection process ols_step_both_adj_r2(model, exclude = c("x2")) # use index of variable instead of name ols_step_both_adj_r2(model, exclude = c(2)) # include & exclude variables in the selection process ols_step_both_adj_r2(model, include = c("x6"), exclude = c("x2")) # use index of variable instead of name ols_step_both_adj_r2(model, include = c(6), exclude = c(2)) ## End(Not run)
Build regression model from a set of candidate predictor variables by entering and removing predictors based on akaike information criteria, in a stepwise manner until there is no variable left to enter or remove any more.
ols_step_both_aic(model, ...) ## Default S3 method: ols_step_both_aic( model, include = NULL, exclude = NULL, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_both_aic' plot(x, print_plot = TRUE, details = TRUE, digits = 3, ...)ols_step_both_aic(model, ...) ## Default S3 method: ols_step_both_aic( model, include = NULL, exclude = NULL, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_both_aic' plot(x, print_plot = TRUE, details = TRUE, digits = 3, ...)
model |
An object of class |
... |
Other arguments. |
include |
Character or numeric vector; variables to be included in selection process. |
exclude |
Character or numeric vector; variables to be excluded from selection process. |
progress |
Logical; if |
details |
Logical; if |
x |
An object of class |
print_plot |
logical; if |
digits |
Number of decimal places to display. |
List containing the following components:
model |
final model; an object of class |
metrics |
selection metrics |
others |
list; info used for plotting and printing |
Venables, W. N. and Ripley, B. D. (2002) Modern Applied Statistics with S. Fourth edition. Springer.
Other both direction selection procedures:
ols_step_both_adj_r2(),
ols_step_both_r2(),
ols_step_both_sbc(),
ols_step_both_sbic()
## Not run: # stepwise regression model <- lm(y ~ ., data = stepdata) ols_step_both_aic(model) # stepwise regression plot model <- lm(y ~ ., data = stepdata) k <- ols_step_both_aic(model) plot(k) # selection metrics k$metrics # final model k$model # include or exclude variables # force variable to be included in selection process model <- lm(y ~ ., data = stepdata) ols_step_both_aic(model, include = c("x6")) # use index of variable instead of name ols_step_both_aic(model, include = c(6)) # force variable to be excluded from selection process ols_step_both_aic(model, exclude = c("x2")) # use index of variable instead of name ols_step_both_aic(model, exclude = c(2)) # include & exclude variables in the selection process ols_step_both_aic(model, include = c("x6"), exclude = c("x2")) # use index of variable instead of name ols_step_both_aic(model, include = c(6), exclude = c(2)) ## End(Not run)## Not run: # stepwise regression model <- lm(y ~ ., data = stepdata) ols_step_both_aic(model) # stepwise regression plot model <- lm(y ~ ., data = stepdata) k <- ols_step_both_aic(model) plot(k) # selection metrics k$metrics # final model k$model # include or exclude variables # force variable to be included in selection process model <- lm(y ~ ., data = stepdata) ols_step_both_aic(model, include = c("x6")) # use index of variable instead of name ols_step_both_aic(model, include = c(6)) # force variable to be excluded from selection process ols_step_both_aic(model, exclude = c("x2")) # use index of variable instead of name ols_step_both_aic(model, exclude = c(2)) # include & exclude variables in the selection process ols_step_both_aic(model, include = c("x6"), exclude = c("x2")) # use index of variable instead of name ols_step_both_aic(model, include = c(6), exclude = c(2)) ## End(Not run)
Build regression model from a set of candidate predictor variables by entering and removing predictors based on p values, in a stepwise manner until there is no variable left to enter or remove any more.
ols_step_both_p(model, ...) ## Default S3 method: ols_step_both_p( model, p_enter = 0.1, p_remove = 0.3, include = NULL, exclude = NULL, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_both_p' plot(x, model = NA, print_plot = TRUE, details = TRUE, ...)ols_step_both_p(model, ...) ## Default S3 method: ols_step_both_p( model, p_enter = 0.1, p_remove = 0.3, include = NULL, exclude = NULL, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_both_p' plot(x, model = NA, print_plot = TRUE, details = TRUE, ...)
model |
An object of class |
... |
Other arguments. |
p_enter |
p value; variables with p value less than |
p_remove |
p value; variables with p more than |
include |
Character or numeric vector; variables to be included in selection process. |
exclude |
Character or numeric vector; variables to be excluded from selection process. |
progress |
Logical; if |
details |
Logical; if |
x |
An object of class |
print_plot |
logical; if |
ols_step_both_p returns an object of class "ols_step_both_p".
An object of class "ols_step_both_p" is a list containing the
following components:
model |
final model; an object of class |
metrics |
selection metrics |
beta_pval |
beta and p values of models in each selection step |
Chatterjee, Samprit and Hadi, Ali. Regression Analysis by Example. 5th ed. N.p.: John Wiley & Sons, 2012. Print.
## Not run: # stepwise regression model <- lm(y ~ ., data = surgical) ols_step_both_p(model) # stepwise regression plot model <- lm(y ~ ., data = surgical) k <- ols_step_both_p(model) plot(k) # selection metrics k$metrics # final model k$model # include or exclude variables model <- lm(y ~ ., data = stepdata) # force variable to be included in selection process ols_step_both_p(model, include = c("x6")) # use index of variable instead of name ols_step_both_p(model, include = c(6)) # force variable to be excluded from selection process ols_step_both_p(model, exclude = c("x1")) # use index of variable instead of name ols_step_both_p(model, exclude = c(1)) ## End(Not run)## Not run: # stepwise regression model <- lm(y ~ ., data = surgical) ols_step_both_p(model) # stepwise regression plot model <- lm(y ~ ., data = surgical) k <- ols_step_both_p(model) plot(k) # selection metrics k$metrics # final model k$model # include or exclude variables model <- lm(y ~ ., data = stepdata) # force variable to be included in selection process ols_step_both_p(model, include = c("x6")) # use index of variable instead of name ols_step_both_p(model, include = c(6)) # force variable to be excluded from selection process ols_step_both_p(model, exclude = c("x1")) # use index of variable instead of name ols_step_both_p(model, exclude = c(1)) ## End(Not run)
Build regression model from a set of candidate predictor variables by entering and removing predictors based on r-squared, in a stepwise manner until there is no variable left to enter or remove any more.
ols_step_both_r2(model, ...) ## Default S3 method: ols_step_both_r2( model, include = NULL, exclude = NULL, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_both_r2' plot(x, print_plot = TRUE, details = TRUE, digits = 3, ...)ols_step_both_r2(model, ...) ## Default S3 method: ols_step_both_r2( model, include = NULL, exclude = NULL, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_both_r2' plot(x, print_plot = TRUE, details = TRUE, digits = 3, ...)
model |
An object of class |
... |
Other arguments. |
include |
Character or numeric vector; variables to be included in selection process. |
exclude |
Character or numeric vector; variables to be excluded from selection process. |
progress |
Logical; if |
details |
Logical; if |
x |
An object of class |
print_plot |
logical; if |
digits |
Number of decimal places to display. |
List containing the following components:
model |
final model; an object of class |
metrics |
selection metrics |
others |
list; info used for plotting and printing |
Venables, W. N. and Ripley, B. D. (2002) Modern Applied Statistics with S. Fourth edition. Springer.
Other both direction selection procedures:
ols_step_both_adj_r2(),
ols_step_both_aic(),
ols_step_both_sbc(),
ols_step_both_sbic()
## Not run: # stepwise regression model <- lm(y ~ ., data = stepdata) ols_step_both_r2(model) # stepwise regression plot model <- lm(y ~ ., data = stepdata) k <- ols_step_both_r2(model) plot(k) # selection metrics k$metrics # final model k$model # include or exclude variables # force variable to be included in selection process model <- lm(y ~ ., data = stepdata) ols_step_both_r2(model, include = c("x6")) # use index of variable instead of name ols_step_both_r2(model, include = c(6)) # force variable to be excluded from selection process ols_step_both_r2(model, exclude = c("x2")) # use index of variable instead of name ols_step_both_r2(model, exclude = c(2)) # include & exclude variables in the selection process ols_step_both_r2(model, include = c("x6"), exclude = c("x2")) # use index of variable instead of name ols_step_both_r2(model, include = c(6), exclude = c(2)) ## End(Not run)## Not run: # stepwise regression model <- lm(y ~ ., data = stepdata) ols_step_both_r2(model) # stepwise regression plot model <- lm(y ~ ., data = stepdata) k <- ols_step_both_r2(model) plot(k) # selection metrics k$metrics # final model k$model # include or exclude variables # force variable to be included in selection process model <- lm(y ~ ., data = stepdata) ols_step_both_r2(model, include = c("x6")) # use index of variable instead of name ols_step_both_r2(model, include = c(6)) # force variable to be excluded from selection process ols_step_both_r2(model, exclude = c("x2")) # use index of variable instead of name ols_step_both_r2(model, exclude = c(2)) # include & exclude variables in the selection process ols_step_both_r2(model, include = c("x6"), exclude = c("x2")) # use index of variable instead of name ols_step_both_r2(model, include = c(6), exclude = c(2)) ## End(Not run)
Build regression model from a set of candidate predictor variables by entering and removing predictors based on schwarz bayesian criterion, in a stepwise manner until there is no variable left to enter or remove any more.
ols_step_both_sbc(model, ...) ## Default S3 method: ols_step_both_sbc( model, include = NULL, exclude = NULL, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_both_sbc' plot(x, print_plot = TRUE, details = TRUE, digits = 3, ...)ols_step_both_sbc(model, ...) ## Default S3 method: ols_step_both_sbc( model, include = NULL, exclude = NULL, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_both_sbc' plot(x, print_plot = TRUE, details = TRUE, digits = 3, ...)
model |
An object of class |
... |
Other arguments. |
include |
Character or numeric vector; variables to be included in selection process. |
exclude |
Character or numeric vector; variables to be excluded from selection process. |
progress |
Logical; if |
details |
Logical; if |
x |
An object of class |
print_plot |
logical; if |
digits |
Number of decimal places to display. |
List containing the following components:
model |
final model; an object of class |
metrics |
selection metrics |
others |
list; info used for plotting and printing |
Venables, W. N. and Ripley, B. D. (2002) Modern Applied Statistics with S. Fourth edition. Springer.
Other both direction selection procedures:
ols_step_both_adj_r2(),
ols_step_both_aic(),
ols_step_both_r2(),
ols_step_both_sbic()
## Not run: # stepwise regression model <- lm(y ~ ., data = stepdata) ols_step_both_sbc(model) # stepwise regression plot model <- lm(y ~ ., data = stepdata) k <- ols_step_both_sbc(model) plot(k) # selection metrics k$metrics # final model k$model # include or exclude variables # force variable to be included in selection process model <- lm(y ~ ., data = stepdata) ols_step_both_sbc(model, include = c("x6")) # use index of variable instead of name ols_step_both_sbc(model, include = c(6)) # force variable to be excluded from selection process ols_step_both_sbc(model, exclude = c("x2")) # use index of variable instead of name ols_step_both_sbc(model, exclude = c(2)) # include & exclude variables in the selection process ols_step_both_sbc(model, include = c("x6"), exclude = c("x2")) # use index of variable instead of name ols_step_both_sbc(model, include = c(6), exclude = c(2)) ## End(Not run)## Not run: # stepwise regression model <- lm(y ~ ., data = stepdata) ols_step_both_sbc(model) # stepwise regression plot model <- lm(y ~ ., data = stepdata) k <- ols_step_both_sbc(model) plot(k) # selection metrics k$metrics # final model k$model # include or exclude variables # force variable to be included in selection process model <- lm(y ~ ., data = stepdata) ols_step_both_sbc(model, include = c("x6")) # use index of variable instead of name ols_step_both_sbc(model, include = c(6)) # force variable to be excluded from selection process ols_step_both_sbc(model, exclude = c("x2")) # use index of variable instead of name ols_step_both_sbc(model, exclude = c(2)) # include & exclude variables in the selection process ols_step_both_sbc(model, include = c("x6"), exclude = c("x2")) # use index of variable instead of name ols_step_both_sbc(model, include = c(6), exclude = c(2)) ## End(Not run)
Build regression model from a set of candidate predictor variables by entering and removing predictors based on sawa bayesian criterion, in a stepwise manner until there is no variable left to enter or remove any more.
ols_step_both_sbic(model, ...) ## Default S3 method: ols_step_both_sbic( model, include = NULL, exclude = NULL, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_both_sbic' plot(x, print_plot = TRUE, details = TRUE, digits = 3, ...)ols_step_both_sbic(model, ...) ## Default S3 method: ols_step_both_sbic( model, include = NULL, exclude = NULL, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_both_sbic' plot(x, print_plot = TRUE, details = TRUE, digits = 3, ...)
model |
An object of class |
... |
Other arguments. |
include |
Character or numeric vector; variables to be included in selection process. |
exclude |
Character or numeric vector; variables to be excluded from selection process. |
progress |
Logical; if |
details |
Logical; if |
x |
An object of class |
print_plot |
logical; if |
digits |
Number of decimal places to display. |
List containing the following components:
model |
final model; an object of class |
metrics |
selection metrics |
others |
list; info used for plotting and printing |
Venables, W. N. and Ripley, B. D. (2002) Modern Applied Statistics with S. Fourth edition. Springer.
Other both direction selection procedures:
ols_step_both_adj_r2(),
ols_step_both_aic(),
ols_step_both_r2(),
ols_step_both_sbc()
## Not run: # stepwise regression model <- lm(y ~ ., data = stepdata) ols_step_both_sbic(model) # stepwise regression plot model <- lm(y ~ ., data = stepdata) k <- ols_step_both_sbic(model) plot(k) # selection metrics k$metrics # final model k$model # include or exclude variables # force variable to be included in selection process model <- lm(y ~ ., data = stepdata) ols_step_both_sbic(model, include = c("x6")) # use index of variable instead of name ols_step_both_sbic(model, include = c(6)) # force variable to be excluded from selection process ols_step_both_sbic(model, exclude = c("x2")) # use index of variable instead of name ols_step_both_sbic(model, exclude = c(2)) # include & exclude variables in the selection process ols_step_both_sbic(model, include = c("x6"), exclude = c("x2")) # use index of variable instead of name ols_step_both_sbic(model, include = c(6), exclude = c(2)) ## End(Not run)## Not run: # stepwise regression model <- lm(y ~ ., data = stepdata) ols_step_both_sbic(model) # stepwise regression plot model <- lm(y ~ ., data = stepdata) k <- ols_step_both_sbic(model) plot(k) # selection metrics k$metrics # final model k$model # include or exclude variables # force variable to be included in selection process model <- lm(y ~ ., data = stepdata) ols_step_both_sbic(model, include = c("x6")) # use index of variable instead of name ols_step_both_sbic(model, include = c(6)) # force variable to be excluded from selection process ols_step_both_sbic(model, exclude = c("x2")) # use index of variable instead of name ols_step_both_sbic(model, exclude = c(2)) # include & exclude variables in the selection process ols_step_both_sbic(model, include = c("x6"), exclude = c("x2")) # use index of variable instead of name ols_step_both_sbic(model, include = c(6), exclude = c(2)) ## End(Not run)
Build regression model from a set of candidate predictor variables by entering predictors based on adjusted r-squared, in a stepwise manner until there is no variable left to enter any more.
ols_step_forward_adj_r2(model, ...) ## Default S3 method: ols_step_forward_adj_r2( model, include = NULL, exclude = NULL, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_forward_adj_r2' plot(x, print_plot = TRUE, details = TRUE, digits = 3, ...)ols_step_forward_adj_r2(model, ...) ## Default S3 method: ols_step_forward_adj_r2( model, include = NULL, exclude = NULL, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_forward_adj_r2' plot(x, print_plot = TRUE, details = TRUE, digits = 3, ...)
model |
An object of class |
... |
Other arguments. |
include |
Character or numeric vector; variables to be included in selection process. |
exclude |
Character or numeric vector; variables to be excluded from selection process. |
progress |
Logical; if |
details |
Logical; if |
x |
An object of class |
print_plot |
logical; if |
digits |
Number of decimal places to display. |
List containing the following components:
model |
final model; an object of class |
metrics |
selection metrics |
others |
list; info used for plotting and printing |
Venables, W. N. and Ripley, B. D. (2002) Modern Applied Statistics with S. Fourth edition. Springer.
Other forward selection procedures:
ols_step_forward_aic(),
ols_step_forward_p(),
ols_step_forward_r2(),
ols_step_forward_sbc(),
ols_step_forward_sbic()
# stepwise forward regression model <- lm(y ~ ., data = surgical) ols_step_forward_adj_r2(model) # stepwise forward regression plot k <- ols_step_forward_adj_r2(model) plot(k) # selection metrics k$metrics # extract final model k$model # include or exclude variables # force variable to be included in selection process ols_step_forward_adj_r2(model, include = c("age")) # use index of variable instead of name ols_step_forward_adj_r2(model, include = c(5)) # force variable to be excluded from selection process ols_step_forward_adj_r2(model, exclude = c("liver_test")) # use index of variable instead of name ols_step_forward_adj_r2(model, exclude = c(4)) # include & exclude variables in the selection process ols_step_forward_adj_r2(model, include = c("age"), exclude = c("liver_test")) # use index of variable instead of name ols_step_forward_adj_r2(model, include = c(5), exclude = c(4))# stepwise forward regression model <- lm(y ~ ., data = surgical) ols_step_forward_adj_r2(model) # stepwise forward regression plot k <- ols_step_forward_adj_r2(model) plot(k) # selection metrics k$metrics # extract final model k$model # include or exclude variables # force variable to be included in selection process ols_step_forward_adj_r2(model, include = c("age")) # use index of variable instead of name ols_step_forward_adj_r2(model, include = c(5)) # force variable to be excluded from selection process ols_step_forward_adj_r2(model, exclude = c("liver_test")) # use index of variable instead of name ols_step_forward_adj_r2(model, exclude = c(4)) # include & exclude variables in the selection process ols_step_forward_adj_r2(model, include = c("age"), exclude = c("liver_test")) # use index of variable instead of name ols_step_forward_adj_r2(model, include = c(5), exclude = c(4))
Build regression model from a set of candidate predictor variables by entering predictors based on akaike information criterion, in a stepwise manner until there is no variable left to enter any more.
ols_step_forward_aic(model, ...) ## Default S3 method: ols_step_forward_aic( model, include = NULL, exclude = NULL, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_forward_aic' plot(x, print_plot = TRUE, details = TRUE, digits = 3, ...)ols_step_forward_aic(model, ...) ## Default S3 method: ols_step_forward_aic( model, include = NULL, exclude = NULL, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_forward_aic' plot(x, print_plot = TRUE, details = TRUE, digits = 3, ...)
model |
An object of class |
... |
Other arguments. |
include |
Character or numeric vector; variables to be included in selection process. |
exclude |
Character or numeric vector; variables to be excluded from selection process. |
progress |
Logical; if |
details |
Logical; if |
x |
An object of class |
print_plot |
logical; if |
digits |
Number of decimal places to display. |
List containing the following components:
model |
final model; an object of class |
metrics |
selection metrics |
others |
list; info used for plotting and printing |
Venables, W. N. and Ripley, B. D. (2002) Modern Applied Statistics with S. Fourth edition. Springer.
Other forward selection procedures:
ols_step_forward_adj_r2(),
ols_step_forward_p(),
ols_step_forward_r2(),
ols_step_forward_sbc(),
ols_step_forward_sbic()
# stepwise forward regression model <- lm(y ~ ., data = surgical) ols_step_forward_aic(model) # stepwise forward regression plot k <- ols_step_forward_aic(model) plot(k) # selection metrics k$metrics # extract final model k$model # include or exclude variables # force variable to be included in selection process ols_step_forward_aic(model, include = c("age")) # use index of variable instead of name ols_step_forward_aic(model, include = c(5)) # force variable to be excluded from selection process ols_step_forward_aic(model, exclude = c("liver_test")) # use index of variable instead of name ols_step_forward_aic(model, exclude = c(4)) # include & exclude variables in the selection process ols_step_forward_aic(model, include = c("age"), exclude = c("liver_test")) # use index of variable instead of name ols_step_forward_aic(model, include = c(5), exclude = c(4))# stepwise forward regression model <- lm(y ~ ., data = surgical) ols_step_forward_aic(model) # stepwise forward regression plot k <- ols_step_forward_aic(model) plot(k) # selection metrics k$metrics # extract final model k$model # include or exclude variables # force variable to be included in selection process ols_step_forward_aic(model, include = c("age")) # use index of variable instead of name ols_step_forward_aic(model, include = c(5)) # force variable to be excluded from selection process ols_step_forward_aic(model, exclude = c("liver_test")) # use index of variable instead of name ols_step_forward_aic(model, exclude = c(4)) # include & exclude variables in the selection process ols_step_forward_aic(model, include = c("age"), exclude = c("liver_test")) # use index of variable instead of name ols_step_forward_aic(model, include = c(5), exclude = c(4))
Build regression model from a set of candidate predictor variables by entering predictors based on p values, in a stepwise manner until there is no variable left to enter any more.
ols_step_forward_p(model, ...) ## Default S3 method: ols_step_forward_p( model, p_val = 0.3, include = NULL, exclude = NULL, hierarchical = FALSE, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_forward_p' plot(x, model = NA, print_plot = TRUE, details = TRUE, ...)ols_step_forward_p(model, ...) ## Default S3 method: ols_step_forward_p( model, p_val = 0.3, include = NULL, exclude = NULL, hierarchical = FALSE, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_forward_p' plot(x, model = NA, print_plot = TRUE, details = TRUE, ...)
model |
An object of class |
... |
Other arguments. |
p_val |
p value; variables with p value less than |
include |
Character or numeric vector; variables to be included in selection process. |
exclude |
Character or numeric vector; variables to be excluded from selection process. |
hierarchical |
Logical; if |
progress |
Logical; if |
details |
Logical; if |
x |
An object of class |
print_plot |
logical; if |
ols_step_forward_p returns an object of class "ols_step_forward_p".
An object of class "ols_step_forward_p" is a list containing the
following components:
model |
final model; an object of class |
metrics |
selection metrics |
Chatterjee, Samprit and Hadi, Ali. Regression Analysis by Example. 5th ed. N.p.: John Wiley & Sons, 2012. Print.
Kutner, MH, Nachtscheim CJ, Neter J and Li W., 2004, Applied Linear Statistical Models (5th edition). Chicago, IL., McGraw Hill/Irwin.
Other forward selection procedures:
ols_step_forward_adj_r2(),
ols_step_forward_aic(),
ols_step_forward_r2(),
ols_step_forward_sbc(),
ols_step_forward_sbic()
# stepwise forward regression model <- lm(y ~ ., data = surgical) ols_step_forward_p(model) # stepwise forward regression plot model <- lm(y ~ ., data = surgical) k <- ols_step_forward_p(model) plot(k) # selection metrics k$metrics # final model k$model # include or exclude variables # force variable to be included in selection process ols_step_forward_p(model, include = c("age", "alc_mod")) # use index of variable instead of name ols_step_forward_p(model, include = c(5, 7)) # force variable to be excluded from selection process ols_step_forward_p(model, exclude = c("pindex")) # use index of variable instead of name ols_step_forward_p(model, exclude = c(2)) # hierarchical selection model <- lm(y ~ bcs + alc_heavy + pindex + enzyme_test, data = surgical) ols_step_forward_p(model, 0.1, hierarchical = TRUE) # plot k <- ols_step_forward_p(model, 0.1, hierarchical = TRUE) plot(k)# stepwise forward regression model <- lm(y ~ ., data = surgical) ols_step_forward_p(model) # stepwise forward regression plot model <- lm(y ~ ., data = surgical) k <- ols_step_forward_p(model) plot(k) # selection metrics k$metrics # final model k$model # include or exclude variables # force variable to be included in selection process ols_step_forward_p(model, include = c("age", "alc_mod")) # use index of variable instead of name ols_step_forward_p(model, include = c(5, 7)) # force variable to be excluded from selection process ols_step_forward_p(model, exclude = c("pindex")) # use index of variable instead of name ols_step_forward_p(model, exclude = c(2)) # hierarchical selection model <- lm(y ~ bcs + alc_heavy + pindex + enzyme_test, data = surgical) ols_step_forward_p(model, 0.1, hierarchical = TRUE) # plot k <- ols_step_forward_p(model, 0.1, hierarchical = TRUE) plot(k)
Build regression model from a set of candidate predictor variables by entering predictors based on r-squared, in a stepwise manner until there is no variable left to enter any more.
ols_step_forward_r2(model, ...) ## Default S3 method: ols_step_forward_r2( model, include = NULL, exclude = NULL, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_forward_r2' plot(x, print_plot = TRUE, details = TRUE, digits = 3, ...)ols_step_forward_r2(model, ...) ## Default S3 method: ols_step_forward_r2( model, include = NULL, exclude = NULL, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_forward_r2' plot(x, print_plot = TRUE, details = TRUE, digits = 3, ...)
model |
An object of class |
... |
Other arguments. |
include |
Character or numeric vector; variables to be included in selection process. |
exclude |
Character or numeric vector; variables to be excluded from selection process. |
progress |
Logical; if |
details |
Logical; if |
x |
An object of class |
print_plot |
logical; if |
digits |
Number of decimal places to display. |
List containing the following components:
model |
final model; an object of class |
metrics |
selection metrics |
others |
list; info used for plotting and printing |
Venables, W. N. and Ripley, B. D. (2002) Modern Applied Statistics with S. Fourth edition. Springer.
Other forward selection procedures:
ols_step_forward_adj_r2(),
ols_step_forward_aic(),
ols_step_forward_p(),
ols_step_forward_sbc(),
ols_step_forward_sbic()
# stepwise forward regression model <- lm(y ~ ., data = surgical) ols_step_forward_r2(model) # stepwise forward regression plot k <- ols_step_forward_r2(model) plot(k) # selection metrics k$metrics # extract final model k$model # include or exclude variables # force variable to be included in selection process ols_step_forward_r2(model, include = c("age")) # use index of variable instead of name ols_step_forward_r2(model, include = c(5)) # force variable to be excluded from selection process ols_step_forward_r2(model, exclude = c("liver_test")) # use index of variable instead of name ols_step_forward_r2(model, exclude = c(4)) # include & exclude variables in the selection process ols_step_forward_r2(model, include = c("age"), exclude = c("liver_test")) # use index of variable instead of name ols_step_forward_r2(model, include = c(5), exclude = c(4))# stepwise forward regression model <- lm(y ~ ., data = surgical) ols_step_forward_r2(model) # stepwise forward regression plot k <- ols_step_forward_r2(model) plot(k) # selection metrics k$metrics # extract final model k$model # include or exclude variables # force variable to be included in selection process ols_step_forward_r2(model, include = c("age")) # use index of variable instead of name ols_step_forward_r2(model, include = c(5)) # force variable to be excluded from selection process ols_step_forward_r2(model, exclude = c("liver_test")) # use index of variable instead of name ols_step_forward_r2(model, exclude = c(4)) # include & exclude variables in the selection process ols_step_forward_r2(model, include = c("age"), exclude = c("liver_test")) # use index of variable instead of name ols_step_forward_r2(model, include = c(5), exclude = c(4))
Build regression model from a set of candidate predictor variables by entering predictors based on schwarz bayesian criterion, in a stepwise manner until there is no variable left to enter any more.
ols_step_forward_sbc(model, ...) ## Default S3 method: ols_step_forward_sbc( model, include = NULL, exclude = NULL, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_forward_sbc' plot(x, print_plot = TRUE, details = TRUE, digits = 3, ...)ols_step_forward_sbc(model, ...) ## Default S3 method: ols_step_forward_sbc( model, include = NULL, exclude = NULL, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_forward_sbc' plot(x, print_plot = TRUE, details = TRUE, digits = 3, ...)
model |
An object of class |
... |
Other arguments. |
include |
Character or numeric vector; variables to be included in selection process. |
exclude |
Character or numeric vector; variables to be excluded from selection process. |
progress |
Logical; if |
details |
Logical; if |
x |
An object of class |
print_plot |
logical; if |
digits |
Number of decimal places to display. |
List containing the following components:
model |
final model; an object of class |
metrics |
selection metrics |
others |
list; info used for plotting and printing |
Venables, W. N. and Ripley, B. D. (2002) Modern Applied Statistics with S. Fourth edition. Springer.
Other forward selection procedures:
ols_step_forward_adj_r2(),
ols_step_forward_aic(),
ols_step_forward_p(),
ols_step_forward_r2(),
ols_step_forward_sbic()
# stepwise forward regression model <- lm(y ~ ., data = surgical) ols_step_forward_sbc(model) # stepwise forward regression plot k <- ols_step_forward_sbc(model) plot(k) # selection metrics k$metrics # extract final model k$model # include or exclude variables # force variable to be included in selection process ols_step_forward_sbc(model, include = c("age")) # use index of variable instead of name ols_step_forward_sbc(model, include = c(5)) # force variable to be excluded from selection process ols_step_forward_sbc(model, exclude = c("liver_test")) # use index of variable instead of name ols_step_forward_sbc(model, exclude = c(4)) # include & exclude variables in the selection process ols_step_forward_sbc(model, include = c("age"), exclude = c("liver_test")) # use index of variable instead of name ols_step_forward_sbc(model, include = c(5), exclude = c(4))# stepwise forward regression model <- lm(y ~ ., data = surgical) ols_step_forward_sbc(model) # stepwise forward regression plot k <- ols_step_forward_sbc(model) plot(k) # selection metrics k$metrics # extract final model k$model # include or exclude variables # force variable to be included in selection process ols_step_forward_sbc(model, include = c("age")) # use index of variable instead of name ols_step_forward_sbc(model, include = c(5)) # force variable to be excluded from selection process ols_step_forward_sbc(model, exclude = c("liver_test")) # use index of variable instead of name ols_step_forward_sbc(model, exclude = c(4)) # include & exclude variables in the selection process ols_step_forward_sbc(model, include = c("age"), exclude = c("liver_test")) # use index of variable instead of name ols_step_forward_sbc(model, include = c(5), exclude = c(4))
Build regression model from a set of candidate predictor variables by entering predictors based on sawa bayesian criterion, in a stepwise manner until there is no variable left to enter any more.
ols_step_forward_sbic(model, ...) ## Default S3 method: ols_step_forward_sbic( model, include = NULL, exclude = NULL, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_forward_sbic' plot(x, print_plot = TRUE, details = TRUE, digits = 3, ...)ols_step_forward_sbic(model, ...) ## Default S3 method: ols_step_forward_sbic( model, include = NULL, exclude = NULL, progress = FALSE, details = FALSE, ... ) ## S3 method for class 'ols_step_forward_sbic' plot(x, print_plot = TRUE, details = TRUE, digits = 3, ...)
model |
An object of class |
... |
Other arguments. |
include |
Character or numeric vector; variables to be included in selection process. |
exclude |
Character or numeric vector; variables to be excluded from selection process. |
progress |
Logical; if |
details |
Logical; if |
x |
An object of class |
print_plot |
logical; if |
digits |
Number of decimal places to display. |
List containing the following components:
model |
final model; an object of class |
metrics |
selection metrics |
others |
list; info used for plotting and printing |
Venables, W. N. and Ripley, B. D. (2002) Modern Applied Statistics with S. Fourth edition. Springer.
Other forward selection procedures:
ols_step_forward_adj_r2(),
ols_step_forward_aic(),
ols_step_forward_p(),
ols_step_forward_r2(),
ols_step_forward_sbc()
# stepwise forward regression model <- lm(y ~ ., data = surgical) ols_step_forward_sbic(model) # stepwise forward regression plot k <- ols_step_forward_sbic(model) plot(k) # selection metrics k$metrics # extract final model k$model # include or exclude variables # force variable to be included in selection process ols_step_forward_sbic(model, include = c("age")) # use index of variable instead of name ols_step_forward_sbic(model, include = c(5)) # force variable to be excluded from selection process ols_step_forward_sbic(model, exclude = c("liver_test")) # use index of variable instead of name ols_step_forward_sbic(model, exclude = c(4)) # include & exclude variables in the selection process ols_step_forward_sbic(model, include = c("age"), exclude = c("liver_test")) # use index of variable instead of name ols_step_forward_sbic(model, include = c(5), exclude = c(4))# stepwise forward regression model <- lm(y ~ ., data = surgical) ols_step_forward_sbic(model) # stepwise forward regression plot k <- ols_step_forward_sbic(model) plot(k) # selection metrics k$metrics # extract final model k$model # include or exclude variables # force variable to be included in selection process ols_step_forward_sbic(model, include = c("age")) # use index of variable instead of name ols_step_forward_sbic(model, include = c(5)) # force variable to be excluded from selection process ols_step_forward_sbic(model, exclude = c("liver_test")) # use index of variable instead of name ols_step_forward_sbic(model, exclude = c(4)) # include & exclude variables in the selection process ols_step_forward_sbic(model, include = c("age"), exclude = c("liver_test")) # use index of variable instead of name ols_step_forward_sbic(model, include = c(5), exclude = c(4))
Test if k samples are from populations with equal variances.
ols_test_bartlett(data, ...) ## Default S3 method: ols_test_bartlett(data, ..., group_var = NULL)ols_test_bartlett(data, ...) ## Default S3 method: ols_test_bartlett(data, ..., group_var = NULL)
data |
A |
... |
Columns in |
group_var |
Grouping variable. |
Bartlett's test is used to test if variances across samples is equal. It is sensitive to departures from normality. The Levene test is an alternative test that is less sensitive to departures from normality.
ols_test_bartlett returns an object of class "ols_test_bartlett".
An object of class "ols_test_bartlett" is a list containing the
following components:
fstat |
f statistic |
pval |
p-value of |
df |
degrees of freedom |
Snedecor, George W. and Cochran, William G. (1989), Statistical Methods, Eighth Edition, Iowa State University Press.
Other heteroskedasticity tests:
ols_test_breusch_pagan(),
ols_test_f(),
ols_test_score()
# using grouping variable if (require("descriptr")) { library(descriptr) ols_test_bartlett(mtcarz, 'mpg', group_var = 'cyl') } # using variables ols_test_bartlett(hsb, 'read', 'write')# using grouping variable if (require("descriptr")) { library(descriptr) ols_test_bartlett(mtcarz, 'mpg', group_var = 'cyl') } # using variables ols_test_bartlett(hsb, 'read', 'write')
Test for constant variance. It assumes that the error terms are normally distributed.
ols_test_breusch_pagan( model, fitted.values = TRUE, rhs = FALSE, multiple = FALSE, p.adj = c("none", "bonferroni", "sidak", "holm"), vars = NA )ols_test_breusch_pagan( model, fitted.values = TRUE, rhs = FALSE, multiple = FALSE, p.adj = c("none", "bonferroni", "sidak", "holm"), vars = NA )
model |
An object of class |
fitted.values |
Logical; if TRUE, use fitted values of regression model. |
rhs |
Logical; if TRUE, specifies that tests for heteroskedasticity be performed for the right-hand-side (explanatory) variables of the fitted regression model. |
multiple |
Logical; if TRUE, specifies that multiple testing be performed. |
p.adj |
Adjustment for p value, the following options are available: bonferroni, holm, sidak and none. |
vars |
Variables to be used for heteroskedasticity test. |
Breusch Pagan Test was introduced by Trevor Breusch and Adrian Pagan in 1979. It is used to test for heteroskedasticity in a linear regression model. It test whether variance of errors from a regression is dependent on the values of a independent variable.
Null Hypothesis: Equal/constant variances
Alternative Hypothesis: Unequal/non-constant variances
Computation
Fit a regression model
Regress the squared residuals from the above model on the independent variables
Compute . It follows a chi square distribution with p -1 degrees of
freedom, where p is the number of independent variables, n is the sample size and
is the coefficient of determination from the regression in step 2.
ols_test_breusch_pagan returns an object of class "ols_test_breusch_pagan".
An object of class "ols_test_breusch_pagan" is a list containing the
following components:
bp |
breusch pagan statistic |
p |
p-value of |
fv |
fitted values of the regression model |
rhs |
names of explanatory variables of fitted regression model |
multiple |
logical value indicating if multiple tests should be performed |
padj |
adjusted p values |
vars |
variables to be used for heteroskedasticity test |
resp |
response variable |
preds |
predictors |
T.S. Breusch & A.R. Pagan (1979), A Simple Test for Heteroscedasticity and Random Coefficient Variation. Econometrica 47, 1287–1294
Cook, R. D.; Weisberg, S. (1983). "Diagnostics for Heteroskedasticity in Regression". Biometrika. 70 (1): 1–10.
Other heteroskedasticity tests:
ols_test_bartlett(),
ols_test_f(),
ols_test_score()
# model model <- lm(mpg ~ disp + hp + wt + drat, data = mtcars) # use fitted values of the model ols_test_breusch_pagan(model) # use independent variables of the model ols_test_breusch_pagan(model, rhs = TRUE) # use independent variables of the model and perform multiple tests ols_test_breusch_pagan(model, rhs = TRUE, multiple = TRUE) # bonferroni p value adjustment ols_test_breusch_pagan(model, rhs = TRUE, multiple = TRUE, p.adj = 'bonferroni') # sidak p value adjustment ols_test_breusch_pagan(model, rhs = TRUE, multiple = TRUE, p.adj = 'sidak') # holm's p value adjustment ols_test_breusch_pagan(model, rhs = TRUE, multiple = TRUE, p.adj = 'holm')# model model <- lm(mpg ~ disp + hp + wt + drat, data = mtcars) # use fitted values of the model ols_test_breusch_pagan(model) # use independent variables of the model ols_test_breusch_pagan(model, rhs = TRUE) # use independent variables of the model and perform multiple tests ols_test_breusch_pagan(model, rhs = TRUE, multiple = TRUE) # bonferroni p value adjustment ols_test_breusch_pagan(model, rhs = TRUE, multiple = TRUE, p.adj = 'bonferroni') # sidak p value adjustment ols_test_breusch_pagan(model, rhs = TRUE, multiple = TRUE, p.adj = 'sidak') # holm's p value adjustment ols_test_breusch_pagan(model, rhs = TRUE, multiple = TRUE, p.adj = 'holm')
Correlation between observed residuals and expected residuals under normality.
ols_test_correlation(model)ols_test_correlation(model)
model |
An object of class |
Correlation between fitted regression model residuals and expected values of residuals.
Other residual diagnostics:
ols_plot_resid_box(),
ols_plot_resid_fit(),
ols_plot_resid_hist(),
ols_plot_resid_qq(),
ols_test_normality()
model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_test_correlation(model)model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_test_correlation(model)
Test for heteroskedasticity under the assumption that the errors are independent and identically distributed (i.i.d.).
ols_test_f(model, fitted_values = TRUE, rhs = FALSE, vars = NULL, ...)ols_test_f(model, fitted_values = TRUE, rhs = FALSE, vars = NULL, ...)
model |
An object of class |
fitted_values |
Logical; if TRUE, use fitted values of regression model. |
rhs |
Logical; if TRUE, specifies that tests for heteroskedasticity be performed for the right-hand-side (explanatory) variables of the fitted regression model. |
vars |
Variables to be used for for heteroskedasticity test. |
... |
Other arguments. |
ols_test_f returns an object of class "ols_test_f".
An object of class "ols_test_f" is a list containing the
following components:
f |
f statistic |
p |
p-value of |
fv |
fitted values of the regression model |
rhs |
names of explanatory variables of fitted regression model |
numdf |
numerator degrees of freedom |
dendf |
denominator degrees of freedom |
vars |
variables to be used for heteroskedasticity test |
resp |
response variable |
preds |
predictors |
Wooldridge, J. M. 2013. Introductory Econometrics: A Modern Approach. 5th ed. Mason, OH: South-Western.
Other heteroskedasticity tests:
ols_test_bartlett(),
ols_test_breusch_pagan(),
ols_test_score()
# model model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) # using fitted values ols_test_f(model) # using all predictors of the model ols_test_f(model, rhs = TRUE) # using fitted values ols_test_f(model, vars = c('disp', 'hp'))# model model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) # using fitted values ols_test_f(model) # using all predictors of the model ols_test_f(model, rhs = TRUE) # using fitted values ols_test_f(model, vars = c('disp', 'hp'))
Test for detecting violation of normality assumption.
ols_test_normality(y, ...) ## S3 method for class 'lm' ols_test_normality(y, ...)ols_test_normality(y, ...) ## S3 method for class 'lm' ols_test_normality(y, ...)
y |
A numeric vector or an object of class |
... |
Other arguments. |
ols_test_normality returns an object of class "ols_test_normality".
An object of class "ols_test_normality" is a list containing the
following components:
kolmogorv |
kolmogorv smirnov statistic |
shapiro |
shapiro wilk statistic |
cramer |
cramer von mises statistic |
anderson |
anderson darling statistic |
Other residual diagnostics:
ols_plot_resid_box(),
ols_plot_resid_fit(),
ols_plot_resid_hist(),
ols_plot_resid_qq(),
ols_test_correlation()
model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_test_normality(model)model <- lm(mpg ~ disp + hp + wt + qsec, data = mtcars) ols_test_normality(model)
Detect outliers using Bonferroni p values.
ols_test_outlier(model, cut_off = 0.05, n_max = 10, ...)ols_test_outlier(model, cut_off = 0.05, n_max = 10, ...)
model |
An object of class |
cut_off |
Bonferroni p-values cut off for reporting observations. |
n_max |
Maximum number of observations to report, default is 10. |
... |
Other arguments. |
# model model <- lm(y ~ ., data = surgical) ols_test_outlier(model)# model model <- lm(y ~ ., data = surgical) ols_test_outlier(model)
Test for heteroskedasticity under the assumption that the errors are independent and identically distributed (i.i.d.).
ols_test_score(model, fitted_values = TRUE, rhs = FALSE, vars = NULL)ols_test_score(model, fitted_values = TRUE, rhs = FALSE, vars = NULL)
model |
An object of class |
fitted_values |
Logical; if TRUE, use fitted values of regression model. |
rhs |
Logical; if TRUE, specifies that tests for heteroskedasticity be performed for the right-hand-side (explanatory) variables of the fitted regression model. |
vars |
Variables to be used for for heteroskedasticity test. |
ols_test_score returns an object of class "ols_test_score".
An object of class "ols_test_score" is a list containing the
following components:
score |
f statistic |
p |
p value of |
df |
degrees of freedom |
fv |
fitted values of the regression model |
rhs |
names of explanatory variables of fitted regression model |
resp |
response variable |
preds |
predictors |
Breusch, T. S. and Pagan, A. R. (1979) A simple test for heteroscedasticity and random coefficient variation. Econometrica 47, 1287–1294.
Cook, R. D. and Weisberg, S. (1983) Diagnostics for heteroscedasticity in regression. Biometrika 70, 1–10.
Koenker, R. 1981. A note on studentizing a test for heteroskedasticity. Journal of Econometrics 17: 107–112.
Other heteroskedasticity tests:
ols_test_bartlett(),
ols_test_breusch_pagan(),
ols_test_f()
# model model <- lm(mpg ~ disp + hp + wt, data = mtcars) # using fitted values of the model ols_test_score(model) # using predictors from the model ols_test_score(model, rhs = TRUE) # specify predictors from the model ols_test_score(model, vars = c('disp', 'wt'))# model model <- lm(mpg ~ disp + hp + wt, data = mtcars) # using fitted values of the model ols_test_score(model) # using predictors from the model ols_test_score(model, rhs = TRUE) # specify predictors from the model ols_test_score(model, vars = c('disp', 'wt'))
Graph to determine whether we should add a new predictor to the model already containing other predictors. The residuals from the model is regressed on the new predictor and if the plot shows non random pattern, you should consider adding the new predictor to the model.
rvsr_plot_shiny(model, data, variable, print_plot = TRUE)rvsr_plot_shiny(model, data, variable, print_plot = TRUE)
model |
An object of class |
data |
A |
variable |
Character; new predictor to be added to the |
print_plot |
logical; if |
model <- lm(mpg ~ disp + hp + wt, data = mtcars) rvsr_plot_shiny(model, mtcars, 'drat')model <- lm(mpg ~ disp + hp + wt, data = mtcars) rvsr_plot_shiny(model, mtcars, 'drat')
Test Data Set
stepdatastepdata
An object of class data.frame with 20000 rows and 7 columns.
A dataset containing data about survival of patients undergoing liver operation.
surgicalsurgical
A data frame with 54 rows and 9 variables:
blood clotting score
prognostic index
enzyme function test score
liver function test score
age, in years
indicator variable for gender (0 = male, 1 = female)
indicator variable for history of alcohol use (0 = None, 1 = Moderate)
indicator variable for history of alcohol use (0 = None, 1 = Heavy)
Survival Time
Kutner, MH, Nachtscheim CJ, Neter J and Li W., 2004, Applied Linear Statistical Models (5th edition). Chicago, IL., McGraw Hill/Irwin.