CODEWITHSUNDEEP
kdb+q-lang
cliff
cliff

Reputation: 43

kdb joining (concatenating) tables with repeating columns

I would like to join (concatenate) 2 or more, KDB tables which have repeating columns:

d1:`a`b`c!(1 2 3;4 5 6;7 8 9) 

d2:`a`b`c`a!(10 20 30;40 50 60;70 80 90;100 110 120)

t1: flip d1 
    a b c
    -----
    1 4 7
    2 5 8
    3 6 9


t2: flip d2
    a  b  c  a  
    ------------
    10 40 70 100
    20 50 80 110
    30 60 90 120

I am trying to concatenate the 2 tables, with nils where a table does not have a value, something like the following:

a  b  c  a  
------------
1  4  7    
2  5  8    
3  6  9    
10 40 70 100
20 50 80 110
30 60 90 120

t1 uj t2 yeilds (only has as many columns as t1 has defined):

a  b  c 
--------
1  4  7 
2  5  8 
3  6  9 
10 40 70
20 50 80
30 60 90

t2 uj t1 (repeats the entries from t1 where they are not available)

a  b  c  a  
------------
10 40 70 100
20 50 80 110
30 60 90 120
1  4  7  1  
2  5  8  2  
3  6  9  3

Is this possible? or would I need to pre-define a table which can accommodate all possible columns? Any advice would be appreciated.

Regards Clifford

Upvotes: 2

Views: 3173

Answers (1)

Rahul
Rahul

Reputation: 3969

Your second table has 2 columns with same name ('a') and that's giving the issue. rename it and uj will work.

q) t1:flip `a`b`c!(1 2 3;4 5 6;7 8 9) 

q) t2:flip `a`b`c`d!(10 20 30;40 50 60;70 80 90;100 110 120)
q) t1 uj t2

 a  b   c   d
 1  4   7   
 2  5   8   
 3  6   9   
 10 40  70  100
 20 50  80  110
 30 60  90  120

Upvotes: 3

Related Questions