|
PureBytes Links
Trading Reference Links
|
I'm not quite sure what you want to have as a final result, but perhaps
this will help.
Use a column for each event. Turn off the filter and run the
exploration. The result set will have the approproate value for the
statement. If there are only 2 events sorting of the columns will
produce usable results. If there are many columns, copy the result set
to Excel and ust the database functions to get to the various subsets
you want.
Ron (not Roy)
Yarroll wrote:
> Hello List,
>
> I have a question which may look silly to most or all of you but the
> more I
> think about it, and especially the combinations of it, the less clear it
> seems to me :-((
>
> I'm trying to use the Explorer to find occurrences when events are
> contained
> within
> the other events. Let's say:
>
> Event1:=C>ref(C,-1);
> Event2:=C>ref(C,-1) and C>mov(C,5,S);
>
> As you can see Event2 is completely contained within the more general
> Event1,
> ie. Event1 is the bigger set here and Event2 is completely contained
> in it,
> or
> there are many cases when Event1 is true while Event2 is not, but not vice
> versa.
>
> This stuff is obvious. But there are indicators and settings not so
> visually
> obvious.
> The right way to explore for these would be to enter the following into
> Explorer:
>
> cum( Event1=0 and Event1=1 )
>
> and this would yield always 0. So far so good.
>
> Important to note here is the sequence here, from the general event to the
> particular.
> What if I don't know which is more general?
> Should I then run the exploration both ways, ie.
> first:
> cum( Event1=0 and Event1=1 )
> and if it yields a number greater than 0, the reverse:
> cum( Event1=1 and Event1=0 )
> and if this one yields greater than 0, then the conclusion would be
> that Event1 and Event2 are either mutually exclusive or overlapping.
>
> But in order to do that I'd have to run 2 explorations, would it be
> possible
> to
> have just one? I have lots of such explorations to run :-((
>
> The only thing I can think of is to first run the exploration to get
> the total number of occurrences of Event1 and Event2, this tells me
> which is
> greater.
>
> Or maybe a formula like:
>
> cum( Event2) / cum( Event1 AND Event2)
>
> if this one yields 1, then we can conclude that Event2 is simply a
> subset of
> Event1.
> Which is exactly what I'm after.
>
> Anything else? Please, even if you have no better idea then please
> acknowledge
> if all this argument makes sense to you at all. I've been thinking of sets
> and
> subsets of events for a while and I'm afraid I'm losing it :-((
>
>
> Thanks, all the best
>
> Yarroll
> ***
> http://republika.pl/yarroll999/
>
>
>
> Yahoo! Groups Sponsor
> ADVERTISEMENT
> <http://rd.yahoo.com/M=243066.2784922.4151385.1927555/D=egroupweb/S=1705001779:HM/A=1377501/R=0/*http://www.verisign.com/cgi-bin/go.cgi?a=b30890113200616000>
>
>
>
> To unsubscribe from this group, send an email to:
> Metastockusers-unsubscribe@xxxxxxxxxxx
>
>
>
> Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service
> <http://docs.yahoo.com/info/terms/>.
To unsubscribe from this group, send an email to:
Metastockusers-unsubscribe@xxxxxxxxxxx
Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
|