Find peers/neighbors amongst the data

[Anonymous]

I have the following data set:

column1, value
ABCD, data1
ABCD, data2

ABCD, data3

ABCD, data4

EFGH, data1

EFGH, data5

EFGH, data6

EFGH, data7

IJKL, data1

IJKL, data2

IJKL, data8

IJKL, data9

...

Q. By load script or an expression how can I determine how similar ABCD is to EFGH and IJKL?

E.g. ABCD has 2 values in common with IJKL: 'data1' and 'data2' . EFGH has 1 value in common with IJKL: 'data1'.

TO RETURN:
column1, peer , count_values_in_common

ABCD,  EFGH, 1                                                 

ABCD , IJKL,  2                                           

EFGH, ABCD, 1           

EFGH, IJKL, 1            

My goal is determine commonalities between one column1 value to another. Thank you in advance.

[Gysbert_Wassenaar · Accepted solution]

See attached qvw. Do not try this on data sets with millions of records.

[Anonymous]

I do have to run this on a rather large number of records, so if you do tweak this script I'd be very keen to see it.

Thank you for this working solution. Very helpful.

[Anonymous]

Gysbert, quick follow-up question. Would you also know how to determine count_values_NOT_in_common?