Using vectors inside dataframe columns in R

Question

I am programming a classificator in R, I have cases that can have multiple classes, so I would like to have both classes in the same row and the column somthing$classes, as a vector. What I mean, is that the column something$classes, of the something data.frame should be like a Java or Python lists of lists. Here is an example of the entering data.frame data:

Case    class    class1     class2
  A       X                    Z
  B                 Y
  C       X         Y          Z
  D                 Y          Z

What I really need to do is to have class, class1 and class2 as one column named classes, with a vector as element, this is the data.frame I would like:

Case    classes  
A       [X, Z] %<- This is a vector, not an string      
B       [Y]          
C       [X, Y, Z]
D       [Y, Z]

Is there a way of having this data.frame structure? If so, how is done and how I could access an individual element inside each classes vector?

Thanks in advance

Data.frames support list columns: DF <- data.frame(a = 1:3); DF$b <- list(1:2, 2:3, 3:4) — Roland
– Roland, Commented Mar 15, 2017 at 8:45
So, I can use something like this in a bucle?: something$classes[i]=list(something$class[i], something$class1[i], something$class2[i]) But this will come in an error if I have some empty elements, like in case A, class1 is empty — Amnor
– Amnor, Commented Mar 15, 2017 at 8:52
Something in base R would be df$classes <- apply(df[-1], 1, function(x) list(x[x!=""])) — Ronak Shah
– Ronak Shah, Commented Mar 15, 2017 at 9:18

akrun · Accepted Answer · 2017-03-15 08:53:16Z

1

We can use data.table

library(data.table)
setDT(df1)[, {v1 <- unlist(.SD); list(classes = list(v1[v1!='']))}, Case]

answered Mar 15, 2017 at 8:53

akrun

891k38 gold badges590 silver badges700 bronze badges

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

7 Comments

Amnor Over a year ago

So, df1 is a new df, or the df I had? I tink I understand that when you do "classes =list(v1[v1!=''])" you are avoyding empty elements, but v1 what exactly is? are you using case as row selector, and the rest of the row should be V1? Thanks!

akrun Over a year ago

@Amnor df1 is the data.frame obect name. Grouped by 'Case', we are unlisting the other columns, remove the blank elements ("") and create a list from the vector elements

Amnor Over a year ago

so df1, is the original object, this instruction, if I want my dataframe to be named "eclides" should be as usual: "euclides <- setDT(df1)[, {v1 <- unlist(.SD); list(classes = list(v1[v1!='']))}, Case]" Am I right? Thanks!

Amnor Over a year ago

Thanks, going for it, just what I needed!!

akrun Over a year ago

It should be euclides$classes[[1]]

|

Collectives™ on Stack Overflow

Using vectors inside dataframe columns in R

1 Answer 1

7 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

7 Comments

Your Answer

Sign up or log in

Post as a guest

Related