I have a table which has 2 columns whose elements are list of symbols. I want to modify one of these columns to append a symbol to existing list of symbols, how can i do it ?
/Table with 2 columns containing list of symbols
tab:([] a:`A`B ;colList1:(`x`y`z;`p`q`r);colList2:(`x`y`z;`p`q`r));
/Gives me a list of symbols in colList2 but not in colList1, I need this for my task
(exec colList1 from tab ) except'(exec colList2 from tab );
/Gives me a list of symbols in colList2 but not in colList1, I need this for my task
(exec colList2 from tab ) except'(exec colList1 from tab );
update missingincol2:(exec colList1 from tab ) except'(exec colList2 from tab ), missingincol1: (exec colList2 from tab ) except'(exec colList1 from tab ) from tab;
update colList1:colList1\:,`ooo from tab / This does not work as intended, I thought for each colList in it will join `ooo to it. But it looks like it create a new column x .
Moreover, when I do meta on a kdb table some of the columns which are list of list gets showed datatype as empty value. If there is a missing column which i want to add in my tickerplant log, how do i do this for a column which is list of list? Note that i do need to modify tp logs and so do not want to use dbmaint script etc. which would make changes directly on hdb level as I replay logs for my task.
Your final update statement should read:
q)update colList1:(colList1,\:`ooo)from tab
a colList1 colList2
--------------------
A x y z ooo x y z
B p q r ooo p q r
Commas indicate a new phrase in the statement so you need brackets to make it explicit that this indicates the join operator. Also the each-left (\:) iterator should be to the right of the operator you're modifying.