|
PureBytes Links
Trading Reference Links
|
Doesn't ColD give the figurative amount of bars since last "formula" result was true?
Lets say that formula was true 238 bars ago, being roughly last Nov97, then
you would need in your filter the smaller then character " < ".
{ and ColD > 240 } should then read as "and ColD < 240"
but you may as well use < 250 since you are loading 300 periods, and
"greater then" would have omited these true happened occurences, the
ones you are looking for.
To keep it happening "about a year ago" you could also add to the filter
that ColD true occurence has to have happened and as such be "greater
then" 200 days ago, resulting in a 50 periods Time Period between 250
and 200 periods ago. This Time Period would then be between Sep97
and Dec97, starting the scan on 981015.
Filter: ColB < 10 and ColD >200 and ColD <250
This set up does restrict the filter to your limitations, and will only bring up
securities as to how you where looking for them, having the criteria:
-stocks which have peaked recently in the last 10 periods(about last 2 weeks)
and had their low point for the year, about a year ago.
But it will not give you:
-all secs having their major low in the past 18 months and then recently, in the
past 6 weeks, peaking.
With Loading min. 400 periods(standard here set to 1300) and
Filter: ColB < 30 and ColD >30 and ColD <350
this will then by pass some of your time limitations and give a somehow slightly
broader time period, so as to not miss out on secs that are left out from having
a strictly applied "year ago" as criteria, but would include secs also complying to your
formula between Apr97 and Aug98, starting this scan on 981015.
Regards,
Ton Maas
Ms-IRB@xxxxxxxxx
-----Oorspronkelijk bericht-----
Van: UG <ug@xxxxxxxxxxxx>
Aan: metastock@xxxxxxxxxxxxx <metastock@xxxxxxxxxxxxx>
CC: support@xxxxxxxxx <support@xxxxxxxxx>
Datum: zondag 18 oktober 1998 15:30
Onderwerp: exploration question
>
>I've got an exploration that just gives me an "N/A" value more often
>than not in one column, and can't quite figure out why.
>
>Here's the nitty-gritty:
>
>ColA: HHV(H, 250)
>ColB: BarsSince(H=HHV(H, 250))
>ColC: LLV(L, 250)
>ColD: BarsSince(L=LLV(L, 250))
>
>Filter: ColB < 10 { and ColD > 240 }
>
>I'm loading 300 periods.
>
>What I'm looking for are stocks which have peaked recently and had their
>low point for the year, about a year ago.
>
>The second part of the filter is commented out because it's the part
>that is giving me trouble. If I leave the filter as-is, I get quite a
>lot of stocks which fit the criteria. When I look at them in the
>Reports screen, ColD (number of days since the LLV), it shows N/A for
>well over 80% of the stocks. I can't find any common thread to why this
>is happening.
>
>Thoughts? Ideas?
>
>--
>========================================================================
>The Pogues have done for Celtic Music what Shane McGowan has done for
>dentistry. -- Roy Gullane, of the Tannahill Weavers
>http://www.unixgeek.com/cgi-bin/motd.pl - PGP email preferred
>
|