cillianreilly
Forum Replies Created
-
You can do this using iteration over a list of the functions to apply.
q){z .(y;x)}[a]/[3;(f2;f1)] 0n 0.5 0.5 0.3333333 0.25 0.2 0.1666667 0.1428571 0.125 0.1111111Adding another function:
q)f3:+ q){z .(y;x)}[a]/[3;(f2;f1;f3)] 0n 1.5 2.5 3.333333 4.25 5.2 6.166667 7.142857 8.125 9.111111 -
You can also include where in the parse tree passed as the last argument.
q)?[tab;til 10;(where;(>;`c2;16))] 7 8 9
-
On your last point, worth mentioning that using brackets to wrap k code is not encouraged: https://code.kx.com/q/basics/exposed-infrastructure/#unary-forms. (I think I heard somewhere that this usage is actually an issue with the parser – but don’t quote me on that.)
-
If you don’t cast to minutes first, you can explicitly group by millisecond (assuming your time column is a timestamp):
q)t:([]time:.z.p+500000*til 20;col:til 20) q)select avg col by 5000000 xbar time from t // millisecond = 1000000 nanoseconds time | col -----------------------------| ---- 2022.09.09D03:50:38.425000000| 2.5 2022.09.09D03:50:38.430000000| 10.5 2022.09.09D03:50:38.435000000| 17.5
-
Another method:
q){a!x a:asc key x:group(!). flip raze key[x],''value x}d 1| ,3 3| 1 2 4| 1 3 5| ,1 6| ,2 7| ,2 -
This achieves what you’re looking for.
q)d:1 2 3!(4 5 3;6 7 3;4 1) q)r:1 3 4 5 6 7!(enlist 3;1 2;1 3;enlist 1;enlist 2;enlist 2) q){a!key[x]where each flip value(a:asc distinct raze x)in/:x} d 1| ,3 3| 1 2 4| 1 3 5| ,1 6| ,2 7| ,2 q)r~{a!key[x]where each flip value(a:asc distinct raze x)in/:x}d 1b -
What is your QHOME variable set to?
C:Users>echo %QHOME%
C:q
You may need to set it to the correct folder location. It is also helpful to add q to your PATH while you are at it:
C:Users>setx QHOME “D:kDBw64”
C:Users>setx PATH “%PATH%;D:kDBw64w64”
Full installation instructions here: https://code.kx.com/q/learn/install
(Being pedantic, you can also remove the top level w64 folder in order to neaten up the environmental variables mentioned.)
-
Yes, this is possible using the system command “d”. See here for more info: https://code.kx.com/q/basics/syscmds/#d-directory
q)d .ns q.ns)fn1:fn2:{}
q)d .
q).ns | :: fn2| {} fn1| {}
-
As David mentioned, you can’t reference iq2 before you create it. If your where clause isn’t too wieldy, you can just include that in the qsql query to create column iq2:
q)select name,iq2:iq%100 from t where 1<iq%100 name iq2 ------------ Prefect 1.26
This does the calculation twice, so be careful with particularly heavy where conditions.
-
The error thrown is passed as the first parameter to the function. https://code.kx.com/q/ref/apply/#trap. To allow parameters, you need to have a function of rank 2 or greater, you can then create a projection of a rank 1 function by eliding the first parameter:
q)logError:{-1"Error: ",x;} q)@[10+;`abc;{logError x,". Couldn't add 10 to ",string y}[;`abc]] Error: type. Couldn't add 10 to abcYou can extend this to allow additional parameters to your error function by creating greater rank functions
-
This is most easily done in two parts and then defaulting the rest to welcome:
q)T:update newCol:`hi from T where a=`F,b<c q)T:update newCol:`hi from T where a<>`F,c>d q)T:update`welcome^newCol from T q)T a b c d newCol ------------------ F 10 9 10 welcome R 20 21 22 welcome
You may find the reference on update statements useful: https://code.kx.com/q/ref/update/
If you want it done in one line (which doesn’t scale well for any additional clauses you might want):
q)update newCol:`welcome`hi any(all(a=`F;b<c);all(a<>`F;c>d))from T a b c d newCol ------------------ F 10 9 10 welcome R 20 21 22 welcome
-
Hi JP, this is just the underlying k implementation of idesc: https://code.kx.com/q/ref/desc/#idesc
q)a:0 1 1 0 -1 0 1 -1 0 0 q)idesc k){$[0h>@x;'`rank;>x]} q)idesc[a]~(>)a 1b q) >a 1 2 6 0 3 5 8 9 4 7The brackets allow q to interpret this, but the use is discouraged (https://code.kx.com/q/basics/exposed-infrastructure/#unary-forms). Rather use the idesc keyword instead.