|
PureBytes Links
Trading Reference Links
|
Thanks Ken for the code.
I have also been working on Multiple time frame testing.....but my code has
been
developed toward the "Gann-Swing" stuff. I have attached a Word.Doc file
with four codes. The first two calculate (on the first day of the week) the
previous
weeks High and Low (also corrected for holidays), the third calculates the
Gann
swing, and the fourth the Gann Weekly Trend.
Best wishes,
Adam Hefner.
VonHef@xxxxxxxxxxxxx
---------------------------------------
----- Original Message -----
From: <divenfish@xxxxxxxxxxxxx>
To: Metastock List <metastock@xxxxxxxxxxxxx>
Sent: Monday, March 29, 1999 10:48 AM
Subject: Weekly moving avg on Daily chart
> Maybe it's the Corona, but I think I'm on to something...<g>
> If I got this right, no matter what day a week started or
> ended on, beginning with the first trading day of each week
> these indicators accurately show the static *weekly* moving
> average on a *daily* chart, making it possible to do multiple
> time frame testing with weekly ma's. (Hey, it's a start..)
> You can
> use the weekly ma's to determine trend, and daily ma's, MACD,
> price...whatever, to enter...all on a daily chart. Let me
> know if you find any errors, but I think it's clean. The only
>
> downside is that it won't plot on the last day loaded since it
>
> needs to look forward one day to determine end of a week.
> One thing I noticed in the past is that sometimes Metastock
> will
> carry the decimal out 3 places on one time frame and sometimes
>
> only two places on another time frame (at least on this
> machine..my
> new one's coming in about a week :-) ), resulting in an
> occasional
> minor (.02% ?) difference between the 50 week ma that's
> plotted on a daily chart and a 50 period ma that's plotted on
> a weekly chart.
>
> Attached are 11 zipped indicator formulae. All will begin
> with ~~ so
> they'll be grouped together.
> The indicator named "~~Wkly SMA's for DAILY Chart" takes a
> LONG time to check things out if you edit the code or rename
> it, so be prepared. It makes 8 fmlvar calls and some are
> quite long. It plots fine speed-wise on the charts,
> considering the number of calcs it has to do. Also, initially
>
> be sure to select only 2,3,4,5,10,20,25, or 50 as the ma
> periods, because those are the only ones I coded it for,
> otherwise it won't plot. If anyone has a suggestion, comment,
> or better way to do it, I'm all ears. I'd like to automate
> some things to eliminate such lenthy code, but at the moment
> nothing comes to mind.
>
> Regards,
> Ken
> --
> mailto:divenfish@xxxxxxxxxxxxx
>
>
Attachment Converted: "c:\eudora\attach\WeeklyTrend.doc"
|