how to solve this error : attempt to make a table with >= 2^31 elements?-CodePudding

I have a data (dt) as below (it is just a simplified example, so you can understand my request):

col 1	col 2	col 3
A	cell 1	1,2
B	cell 2	3,5,6
A	cell 3	1,2
D	cell 4	9,7,8

What I want to know is in how many cells I can have the same combinations of col1 and col3 :

So the output should be like this :

col 1	col 3	col 2
A	1,2	2 cells
B	3,5,6	1 cell
D	9,7,8	1 cell

I was trying this command using "data.table" in r :

data.table(table(dt$col1,dt$col3))

However it gives me this error :

error in table(dt$col1,dt$col3) : attempt to make a table with >= 2^31 elements

I am doing this in R, data.table

CodePudding user response：

You can just aggregate, i.e.

library(data.table)
setDT(d2)[, .(col2 = .N, col3 = toString(unique(col3))), col1][]

#   col1 col2  col3
#1:    A    2   1_2
#2:    B    1 3_5_6
#3:    D    1 9_7_8

CodePudding user response：

Try this

library(dplyr)

df %>% group_by(col1 , col3) %>% summarise("coll2" = paste(length(col3) , "cells"))

CodePudding user response：

We could use count from dplyr and its name argument:

df %>% 
  count(col1, col3, name = "col2")

  col1  col3 col2
1    A   1,2    2
2    B 3,5,6    1
3    D 9,7,8    1

In case you need cell as string, we could add mutate(col2 = paste(col2, "cell"))to the code.