Luigi,

thanks. I am a bit reluctant to share the query itself here at this time. 
However, your answer actually reminded me of another
issue. Although the match with 'where: ...'  works as expected, returning 
all data (very fast also when limit is 400+). 
Applying my 'arcsinh(arg)' function to columns in the return part did not 
work. Hence the attempt  with select, which works, but very very slow even 
when adding PARALLEL(which even in my 2.2.5 makes a big difference).

By the way, how can MATCH show data from two "branches" (or a fork) from a 
vertex with two edges? 
  
As you say, MATCH is probably not yet fully matured, so I am looking 
forward to upcoming versions. Having said this, I would like to add
that in the latest releases you made fantastic progress!   

Regarding the alias part in the RETURN section. From my point of view, the 
example I referred from you blog post is perhaps a bit
superfluous? Why return aliases when the select columns would need yet 
another set of aliases as well? 
 
T.

On Thursday, July 21, 2016 at 9:10:41 PM UTC+2, Tore wrote:
>
> 1) Match works beautifully(once I got the hang of it). However as 
> suggested in the otherwise excellent blog post, "Pattern matching with 
> OrientDB",
>     select something from (match .....) takes ages while the match itself 
> (i.e. without select) returns in milliseconds. How come?
>
> 2) In the same example(above) select from (match   ,) the suggested return 
> of aliases, as RETURN A as a, B as b;  do not work (2.2.4).
>
> Finally - a tiny detail at towards the very end of in the same blog, 
> "Remember always have to use RETURN $elements or RETURN $pathElements".
>
> *Thank you Luigi! *That really saved my day (just like the day when I 
> discovered . in vi, to repeat last command) Hurray!
>
>
>

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"OrientDB" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to