[R] How to run a model 1000 times, while saving coefficients each time?

[Mark Na]

Hello,

We have written a program (below) to model the effect of a covariate on 
observed values of a response variable (using only 80% of the rows in 
our dataframe) and then use that model to calculate predicted values for 
the remaining 20% of the rows. Then, we compare the observed vs. 
predicted values using a linear model and inspect that model's 
coefficients and its R2 value.

We wish to run this program 1000 times, and to save the coefficients and 
R2 values into a separate dataframe called results.

We have a looping structure (also below) but we do not know how to save 
the coefficients and R2 values. We are missing some code (indicated)

Any assistance would be greatly appreciated.

Thanks,


library(sampling)

mall<-read.csv("mall.csv")

for (j in 1:1000) {

s<-srswor(2840,3550)
mall80<-mall[s==1,]
mall20<-mall[s==0,]
model1<-lm(count~habitat,data=mall80)
summary(model1)
mall20$predicted<-predict(model1,newdata=mall20)
model2<-lm(count~predicted,data=mall20)

MISSING CODE: SAVE MODEL COEFFICIENTS AND R2 VALUE TO A DATAFRAME CALLED 
RESULTS

}