Fit regression lines of 500 samples of N=50 onto one sample plot in R?-CodePudding

I need to get 500 samples of N=50 each and draw a simple linear regression model with X1 as VI and y as DV on each sample. Then, I have to plot all of these samples' predictions on one previous sample dataframe, which I've called sample.dat.

This is my code so far:

  geom_point(sample.dat, mapping = aes(x = X1, y = y))  
  #geom_line(data = lm.fit.dat, aes(y = true.y.fit), color = "blue")  
  theme_bw()


    for(i in 1:500){
  df = linear.dat[sample(1:nrow(linear.dat), size = 50),]
  g = p   geom_smooth(method = lm, data=df, color="red", size=0.5, alpha = 0)  
  geom_line(data = lm.fit.dat, aes(y = true.y.fit), color = "blue")  
}
    plot(g)

And this is my output: Output

As you see, I only have 1 red line when I want 500 red lines of the 500 samples.

CodePudding user response：

The big problem is that in the loop you're adding the new line to p, but g isn't accumulating the lines. At the end of the loop, g will only have the most recently generated line. To accumulate the lines you have to do g <- g ... - you have to add the current line to all of the previous lines. Something like this should work:

g <- ggplot(sample.dat, mapping = aes(x = X1, y = y))   
  geom_point()  
  theme_bw()


for(i in 1:500){
  df = linear.dat[sample(1:nrow(linear.dat), size = 50),]
  g = g   
    geom_smooth(method = lm, data=df, color="red", size=0.5, alpha = 0) 

}
plot(g)