What are the FREE and easy to use software packages that available to compare the appropriateness of a number of regression models without writing scripts/commands? For example, STATGRAPHICS (not Free though) allows to analyse around 15 models for simple regression based on R squared. What else we have?
Thanks!
Get a copy of Jared Lander R for everyone from the z-library. Details and programs are included. Just copy them and get started. Best wishes, David Booth
In what way do you want to compare models ? ... I do have an R function that compares general linear models for AIC, AICc, BIC, R.squared, Adj.R.sq, and p.value . The following example can be run at https://rdrr.io/snippets/ .
#############
### Compare among polynomial models
library(rcompanion)
data(BrendonSmall)
BrendonSmall$Calories = as.numeric(BrendonSmall$Calories)
BrendonSmall$Calories2 = BrendonSmall$Calories * BrendonSmall$Calories
BrendonSmall$Calories3 = BrendonSmall$Calories * BrendonSmall$Calories *
BrendonSmall$Calories
BrendonSmall$Calories4 = BrendonSmall$Calories * BrendonSmall$Calories *
BrendonSmall$Calories * BrendonSmall$Calories
model.1 = lm(Sodium ~ Calories, data = BrendonSmall)
model.2 = lm(Sodium ~ Calories + Calories2, data = BrendonSmall)
model.3 = lm(Sodium ~ Calories + Calories2 + Calories3, data = BrendonSmall)
model.4 = lm(Sodium ~ Calories + Calories2 + Calories3 + Calories4,
data = BrendonSmall)
compareLM(model.1, model.2, model.3, model.4)
##### Results #####
# $Models
# Formula
# 1 "Sodium ~ Calories"
# 2 "Sodium ~ Calories + Calories2"
# 3 "Sodium ~ Calories + Calories2 + Calories3"
# 4 "Sodium ~ Calories + Calories2 + Calories3 + Calories4"
#
# $Fit.criteria
# Rank Df.res AIC AICc BIC R.squared Adj.R.sq p.value Shapiro.W Shapiro.p
# 1 2 43 461.1 461.7 466.5 0.7207 0.7142 1.737e-13 0.9696 0.2800
# 2 3 42 431.4 432.4 438.6 0.8621 0.8556 8.487e-19 0.9791 0.5839
# 3 4 41 433.2 434.7 442.2 0.8627 0.8526 1.025e-17 0.9749 0.4288
# 4 5 40 428.5 430.8 439.4 0.8815 0.8696 5.566e-18 0.9625 0.1518
the performance package might be useful https://easystats.github.io/performance/
Manjula Wijewickrema -
You mentioned comparing models using R-squared, but that is not a good idea. See https://data.library.virginia.edu/is-r-squared-useless/. There they note that R-squared is not a measure of fit. It is often misinterpreted, but I think there that they meant that because the model part of the variance can be low and the sigma part can be high, you could have a low R-squared, but still the best model for the application. At any rate, I don't think one statistic tells you that much. I suggest the graphical approach. For a given sample, you can compare model fits by using a "graphical residual analysis." To avoid fitting to your sample too closely, such that your model does not predict well for the remainder of the subpopulation or population you wish to model, considering a "cross-validation." You could research those two terms. If you just try to have an automatic comparison between models without really knowing anything about them, you may not really get what you want.
Cheers - Jim