Using R + For Loops to Create Variables

Question

Supposed that I have the following data:

a1 <- 1
a2 <- 2
a3 <- 3

In my actual data, I have way more variable and what I am aiming to do is create interaction variables between them and then put them into one big dataframe like this:

a1a1 <- a1*a1
a1a2 <- a1*a2
a1a3 <- a1*a3

a2a2 <- a2*a1
a2a3 <- a2*a2

a3a3 <- a3*a3

mydf <- data.frame(a1a1,a1a2,...)

I am certain I could use a forloop to accomplish what I want but I am horribly new to R and coding methodology altogether. My intuition tells me to do something along the lines of this:

 n <- 3
 j <- 3

 for( i in 1:n) {
    k <- 1
    for( j in 1:k) {

 aiaj <- ai*aj 
 #Not sure how to put into a data.frame for each variable
 k <- k + 1 
 }
 }

The way the loop is constructed, I believe it would make the variables a1a1,a1a2,a1a3 on the first iteration, on the second iteration it would make a2a2,a2a3, and then on the third iteration a3a3. I've tried this, and it doesn't work so I am hoping I can get some suggestions. Additionally, I have no idea how I would keep adding to the data.frame

Do you really need to create those variables? Many R functions can handle these interactions via a formula. — G5W
– G5W, Commented Apr 1, 2017 at 0:23
Perhaps I do not, I don't really know. But, I do know that if I do create these variables, then I will be able to do what I need to do for the purpose of the original data. — Dennis Gray
– Dennis Gray, Commented Apr 1, 2017 at 0:36
For example, If you are trying to generate a linear model with these cross-terms you can use a formula like lm(y ~ (a1+a2+a3)^2, data=data). If the function that you are trying to use accepts formulas. please take a look at the help page ?formula — G5W
– G5W, Commented Apr 1, 2017 at 0:41
What is the different between a1*a2 and a2*a1? Moreover, why a2*a3 is missing? — YQ.Wang
– YQ.Wang, Commented Apr 1, 2017 at 0:45
I corrected the difference. There is no difference between a1*a2, and I added a2*a3. They should all be unique. — Dennis Gray
– Dennis Gray, Commented Apr 1, 2017 at 1:35

manotheshark · Accepted Answer · 2017-04-01 02:12:49Z

2

This will store all interactions in a data.frame that can filtered or additional functions used as desired

a <- c(a1, a2, a3)
df1 <- expand.grid(a, a)
df1$prod <- df1[, 1] * df1[, 2]

> df1
  Var1 Var2 prod
1    1    1    1
2    2    1    2
3    3    1    3
4    1    2    2
5    2    2    4
6    3    2    6
7    1    3    3
8    2    3    6
9    3    3    9

answered Apr 1, 2017 at 2:12

manotheshark

4,36520 silver badges31 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

Kay · Accepted Answer · 2017-04-01 01:50:55Z

0

How about you try this:

suppose this is your data frame

q1 <- data.frame(a = sample(1:20, 4), b = sample(1:20, 4),c=sample(1:20, 4))

you can use lapply to interact the columns with this function. There may be efficient methods but I hope this solves your problem.

library(dplyr)

 mn <- lapply(1:3, function(j) data.frame(lapply(1:3, function(i) 
         q1[,i]*q1[,j])))%>%bind_cols()

Column names may not be pretty but you can change them at any time using colnames

edited Apr 1, 2017 at 1:50

answered Apr 1, 2017 at 0:45

Kay

2,3825 gold badges24 silver badges32 bronze badges

Collectives™ on Stack Overflow

Using R + For Loops to Create Variables

2 Answers 2

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related