blr_step_p_backward: Stepwise backward regression

In blorr: Tools for Developing Binary Logistic Regression Models

Description

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.

Usage

blr_step_p_backward(model, ...)

## Default S3 method:
blr_step_p_backward(model, prem = 0.3, details = FALSE, ...)

## S3 method for class 'blr_step_p_backward'
plot(x, model = NA, print_plot = TRUE, ...)

Arguments

model	An object of class lm; the model should include all candidate predictor variables.
...	Other inputs.
prem	p value; variables with p more than prem will be removed from the model.
details	Logical; if TRUE, will print the regression result at each step.
x	An object of class blr_step_p_backward.
print_plot	logical; if TRUE, prints the plot else returns a plot object.

Value

blr_step_p_backward returns an object of class "blr_step_p_backward". An object of class "blr_step_p_backward" is a list containing the following components:

model	model with the least AIC; an object of class glm
steps	total number of steps
removed	variables removed from the model
aic	akaike information criteria
bic	bayesian information criteria
dev	deviance
indvar	predictors

References

Chatterjee, Samprit and Hadi, Ali. Regression Analysis by Example. 5th ed. N.p.: John Wiley & Sons, 2012. Print.

See Also

Other variable selection procedures: blr_step_aic_backward(), blr_step_aic_both(), blr_step_aic_forward(), blr_step_p_forward()

Examples

## Not run: 
# stepwise backward regression
model <- glm(honcomp ~ female + read + science + math + prog + socst,
  data = hsb2, family = binomial(link = 'logit'))
blr_step_p_backward(model)

# stepwise backward regression plot
model <- glm(honcomp ~ female + read + science + math + prog + socst,
  data = hsb2, family = binomial(link = 'logit'))
k <- blr_step_p_backward(model)
plot(k)

# final model
k$model


## End(Not run)

blorr documentation built on July 2, 2020, 2:15 a.m.