|
PureBytes Links
Trading Reference Links
|
<SPAN
class=395093920-03022004>Thanks for your further investigations, good work.
Interesting that it's only direct uses of the reserved variables that
causes this. Besides your stack overflow idea, maybe there's an issue with
how input parameters are handled. I'd think stack issues would affect any
repeated call to the function, regardless of what the input vars were, but of
course I don't have much context for even guessing.
<SPAN
class=395093920-03022004>
<SPAN
class=395093920-03022004>I did write Tomasz, haven't heard back, we'll see what
he says.
<SPAN
class=395093920-03022004>
<SPAN
class=395093920-03022004>Thanks again for checking, glad to know I'm not just
completely losing it over here. I think...
<SPAN
class=395093920-03022004>
<SPAN
class=395093920-03022004>Dave
<BLOCKQUOTE
>I
went a little further. I'd encourage you to send it to TJ. I
think that it may be a serious problem. I instrumented the code and
played with it a little. Using a ticker of DJ-30, my conclusion was
that a call to StochTransform with any of the reserved variables (O,
C, etc.) overwrote the variable that was used in the previous call.
Note in the example how StochDiff is overwritten. In traditional
languages, this type of thing is usually indicative of a stack problem
in the function processing.Here's what I played with -function
StochTransform(array, period) { local
LowValue; LowValue = LLV(array,
period); return 100 * (array - LowValue) /
(HHV(array, period) - LowValue);}Filter = 1;Diff =
Signal(12, 26, 9);_TRACE(WriteVal(Diff, 8.3));AddColumn(Diff,
"DIFF BEFORE", 8.3);StochDiff = StochTransform(Diff,
5);// It looks like a call to StochDiff with Close as the first
parm wipes// out the variable that was used in the last call to
StochDiffJunkIgnored = StochTransform(Diff, 14); // <<< BUT THIS
DOESN'TJunkIgnored = StochTransform(StochDiff, 14); // <<< AND
THIS DOESN'T//JunkIgnored = StochTransform(C + 0, 14); // <<< AND
THIS DOESN'TJunkIgnored = StochTransform(C, 14); // <<< ENABLING
THIS CHANGES Diff,WHY???_TRACE(WriteVal(Diff,
8.3));AddColumn(Close, "CLOSE", 8.3);AddColumn(Diff, "DIFF AFTER",
8.3);AddColumn(StochDiff, "STOCHDIFF", 8.3);AddColumn(JunkIgnored,
"JUNK", 8.3);-- Bruce--- In amibroker@xxxxxxxxxxxxxxx,
"Dave Merrill" <dmerrill@xxxx> wrote:> Thanks for checking it
out Bruce, I appreciate it. The msg I posted was a> simplification
of the more complicated code where I first found the problem,> and
that was the last work I've had time to do on it. (That j.o.b.
thing...)> > Do I gather that you saw the same results I
did? Puzzling, isn't it?> > Dave> I was up
early this morning and took at quick look at the problem>
that you found. Have you made any headway?> >
-- Bruce> > > --- In
amibroker@xxxxxxxxxxxxxxx, "Dave Merrill"
<dmerrill@xxxx>> wrote:> > I'm
seeing a weird problem that I think may be the result of>
something I> > don't understand about variable scoping.
Check this out:> >> >
----------------------> > function StochTransform(array,
period) {> > local
LowValue;> >
LowValue = LLV(array, period);>
> return 100 * (array - LowValue) /
(HHV(array, period) -> LowValue);> >
}> >> > Diff = Signal(12, 26,
9);> > StochDiff = StochTransform(Diff,
5);> >> > //JunkIgnored =
StochTransform(C, 14); // <<< ENABLING
THIS> CHANGES Diff,> >
WHY???> > //JunkIgnored = StochTransform(Diff,
14); // <<< BUT
THIS> DOESN'T> >
----------------------> >> > First
enter it into IB as given above, look at the chart, then>
enable the> > first of the two commented out lines at
the bottom. Notice that> Diff (blue>
> histogram) changes drastically. Now enable the second one and
not> the first,> > and things go
back to "normal".> >> > I don't
get this at all. It seems like it must be due to>
interaction between> > the two calls to StochTransform,
and in fact, if you create a copy> of
the> > function called StochTransform2, and call that
the second time, no> problem.> >
Guessing further from the fact that using the same value
for> the 'array'> > parameter in both
cases also prevents the problem, most likely the>
conflict> > is with the reuse of that formal parameter
when the function is> called the> >
second time.> >> > I've never seen
this behavior before, and I don't understand it.> Among
other> > things, if it's really true that you can't call
the same function> more than> >
once without interactions like this, a lot of code I've
written> that I> > thought worked,
didn't. Or is there some simple stupid bug in this>
test that> > I'm missing?>
>> > Ideas? Tomasz?>
>> > Dave
Send BUG REPORTS to bugs@xxxxxxxxxxxxx
Send SUGGESTIONS to suggest@xxxxxxxxxxxxx
-----------------------------------------
Post AmiQuote-related messages ONLY to: amiquote@xxxxxxxxxxxxxxx
(Web page: http://groups.yahoo.com/group/amiquote/messages/)
--------------------------------------------
Check group FAQ at: http://groups.yahoo.com/group/amibroker/files/groupfaq.html
Yahoo! Groups Links
To visit your group on the web, go to:http://groups.yahoo.com/group/amibroker/
To unsubscribe from this group, send an email to:amibroker-unsubscribe@xxxxxxxxxxxxxxx
Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.
|