Create an adjacency from dataframe with two columns with R

Question

I have a dataframe that has this structure

ID LINEITEM
1  shirt1
1  shirt2
2  shirt1
3  shirt3
3  shirt4
4  shirt5
5  shirt1
5  shirt2
5  shirt6

How can I shape the data in order to create edges and nodes between the LINEITEMS according to how many times they appear on the same ID?

My expected output would be:

       Shirt1 Shirt2 Shirt3 Shirt4 Shirt5 Shirt6
Shirt1   0      2       0       0      0       1   
Shirt2   2      0       0       0      0       0   
Shirt3   0      0       0       1      0       0   
Shirt4   0      0       1       0      0       0   
Shirt5   0      0       0       0      0       0   
Shirt6   1      1       0       0      0       0

which counts every time two shirts appear on a same ID together)

3

what is your expected output ? – BENY Jul 27 '19 at 00:14

BENY · Answer 1 · 2019-07-27T00:48:12.737

3

Check with table

table(df$ID, df$LINEITEM)

As r2evans mentioned in comments

xtabs(~ ID + LINEITEM, data = df)

edited Jul 27 '19 at 00:48

answered Jul 27 '19 at 00:38

BENY

317,841
20
164
234

2

Similarly, `xtabs(~ ID + LINEITEM, data = df)` – r2evans Jul 27 '19 at 00:39
@r2evans would you mind I adding that :-) – BENY Jul 27 '19 at 00:40
1

Feel free to add it, they go together (the biggest difference I find is that `xtabs` goes beyond two dims, but sometimes `table` is sufficient for me). – r2evans Jul 27 '19 at 00:47

Create an adjacency from dataframe with two columns with R

1 Answers1