How to declare a for loop in R

Question

I am trying to create a simple for loop in R, but I am not sure how to go about this without creating a global variable.

I am trying to output a predict table neatly, instead of running code through many different instances (something like below) that I wish to predict.

house1 = newdata[1,]
predict(fullmodel, house1)
predict(sqftmodel, house1)
predict(bestmodel, house1)
house2 = newdata[2,]
predict(fullmodel, house2)
predict(sqftmodel, house2)
predict(bestmodel, house2)
house3 = newdata[3,]

I want to use a for loop to run through 37 different houses and have the output in a table. Any ideas?

edit: this is a portion of my code so far

data = read.table("DewittData.txt")
 newdata =  na.omit(data)#28 points to refer to
colnames(newdata) = c("ListPrice", "Beds", "Bath","HouseSize","YearBuilt"
         ,"LotSize", "Fuel","ForcedAir", "Other","FM","ESM","JD",
        "SchoolDistrict","HouseType","GarageStalls","Taxes")
attach(newdata)
fullmodel = lm((ListPrice) ~ HouseSize + Beds + Bath + YearBuilt + LotSize
                + Fuel + ForcedAir + Other + SchoolDistrict+
                HouseType + Other + FM + ESM + JD + GarageStalls + Taxes)

bestmodel = lm(ListPrice~Beds)
sqftmodel = lm(ListPrice~HouseSize, data = newdata)

update:

I see, so I've changed it to

predict(fullmodel, newdata[,])
predict(sqftmodel, newdata[,])
predict(bestmodel, newdata[,])

Now how would I output this in a table format?

One way to solve your question is to build your function. Or may use 'apply' function. — , May 03 '17 at 15:56
You can pass the whole dataframe to `predict()` directly, no need to pass rows one by one — Aurèle, May 03 '17 at 16:03
@BeginnerJava Use `dput` to make a reproducible example of your data set. — M--, May 03 '17 at 16:11
@BeginnerJava Sorry I should've included this link: http://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example — M--, May 03 '17 at 16:16
I've got what i need, but I'm not sure how to output in in table without manually entering data — Beginner Java, May 03 '17 at 16:26

score 1 · Accepted Answer · answered May 03 '17 at 16:09

1

I am not sure if I get your question , but this what I would do for predicting based on different rows of a df.

Housefull <-  predict(fullmodel, newdata[,])
Housebest <-  predict(bestmodel, newdata[,])
Housesqft <-  predict(sqftmodel, newdata[,])

Generally, sticking to vectors is much better than using loops.

answered May 03 '17 at 16:09

M--

25,431
8
61
93

As I understand you, you want to do all your code at once. So, I built my function to you but since I cannot access your data I get an error. I will pass you my code and just replace your code in and see if it is work or not. – May 03 '17 at 16:13

How to declare a for loop in R

1 Answers1