How to plot results from analysis of different datasets into one graph I conducted the same logistic regression analysis for five different datasets. Now I would like to plot the results (the Odds-Ratios) all into one graph. Is this even possible? And how would I do it? 
 A: What to plot and how to plot are depending on the story / insights you want to tell.
I can think about many ways of showing 5 different models in one plot. But still do not know your goal. Here are two examples:


*

*Plot five logistic curves on 5 data set, i.e, you will see 5 'S-shaped' curve, where x is the log odds and y is the probability. The goal here is you can compare the 'bias' term in different models, i.e, the red curve is centered at 1 but the green curve is centered at -1. Also you can compare the how steep / flat of different models.





*

*Use histogram (such as stacked bar chart) to show output probably distribution of 5 models.

A: Using the mtcars data:
mod1<- glm(vs~hp, data = mtcars, family = 'binomial')
mod2 <- glm(vs~disp, data = mtcars, family = 'binomial')
mod3 <- glm(vs~cyl, data = mtcars, family = 'binomial')


#Creating predictor dataset for each of the variables of length 100

newdata <- data.frame(hp=seq(min(mtcars$hp), max(mtcars$hp),len=100),
                      disp=seq(min(mtcars$disp), max(mtcars$disp), len=100),
                      cyl=seq(min(mtcars$cyl), max(mtcars$cyl), len=100))

#Predicting the dependent variables using each model
newdata$vs1 <- predict(mod1, newdata=newdata, type="response")
newdata$vs2 <- predict(mod2, newdata=newdata, type="response")
newdata$vs3 <- predict(mod3, newdata=newdata, type="response")

newdata[1:3]<- scale(newdata[1:3])  #scaling 3 predictor variables for standardizing x axis

plot(vs1~hp, data=newdata, type='l', col="red", xlab="Predictor (Scaled)", ylab="Predicted", lwd =2)
par(mfrow=c(1,1), new=TRUE)
plot(vs2~disp, data=newdata, type='l', col="blue", xaxt='n', yaxt='n', xlab="" , ylab="", lwd =2)
par(mfrow=c(1,1), new=TRUE)
plot(vs3~cyl, data=newdata, type='l', col="darkgreen", xaxt='n', yaxt='n',  xlab="" , ylab="", lwd =2)


