How do you loop through a regression?

[glittergirl]

I am trying to loop through a regression but am having issues. Here is my reproducible data:

data(iris)

  model1 <- lm(Sepal.Length ~ Sepal.Width, data=iris1)
  model2 <- lm(Sepal.Length ~ Sepal.Width + Petal.Length + Petal.Width, data=iris1)

predict1 <- function(value, model){ 
  y = model$coefficients[1] + (model$coefficients[-1]%*% value)
  return (y)
}

predict2 <- function(values, model){
  y = model$coefficients[1] + (model$coefficients[-1]%*% values)
  return (y)
}

n = 300

for (i in 1:n){
  
  iris1 <- iris[sample(1:nrow(iris), 50, replace = TRUE),]

## I suspect there is something wrong here, but I am not sure what the issue is...
  
  model1[[i]] <- lm(Sepal.Length ~ Sepal.Width, data=iris1)
  model2[[i]] <- lm(Sepal.Length ~ Sepal.Width + Petal.Length + Petal.Width, data=iris1)
  
  predictions1 <- predict1(iris1$Sepal.Width[[i]], model1[[i]])
  values <- c(iris1$Sepal.Width, iris1$Petal.Length, iris1$Petal.Width)
  predictions2 <- predict2(values[[i]], model2[[i]])
}

[technocrat]

Great reprex!

Two things pop out.

n <- ????

model1[[i]] <- ... modifies model1, but it doesn't exist.

[glittergirl]

I just edited the code and defined n and model1 and model2.

[andresrcs]

Sorry but your code doesn't make sense and is not clear what you are trying to accomplish. Is this close to what you are trying to do?

n = 300
predictions1 <- double()
predictions2 <- double()

for (i in 1:n){
    
    iris1 <- iris[sample(1:nrow(iris), 50, replace = TRUE),]
    
    model1 <- lm(Sepal.Length ~ Sepal.Width, data=iris1)
    model2 <- lm(Sepal.Length ~ Sepal.Width + Petal.Length + Petal.Width, data=iris1)
    
    predictions1 <- c(predictions1, predict(model1, newdata = iris1))
    predictions2 <- c(predictions2, predict(model2, newdata = iris1))
}