|
PureBytes Links
Trading Reference Links
|
Craig, Guy &all,
The not available function for a variable (thus replaceable) variable, is indeed a cumbersome
problem in MS.
However, if you are enabled to manipulate, you are also enabled to repair (eg un-manipulate
to the original state).
What I mean is that Ref(U1,-1) is only here, yesterday's manipulated and -adapted from- outcome
value (the "cocktail"), of some kind of "base" value (the "basic").
Thus by creating a custom indicator of this "base" value, eg the basic indicator that is not
manipulated to the U1 extend, would than provide for the required daily source for "the other" bucket
values, and these "otherwise perhaps manipulated" daily values can now also be used for a single
seperate indicator (eg "The Bucket-indicator").
Example-A.
Where the Close can be a calculated for a single separate custom Price-indicator (eg C),
the Bucket-indicator can also be a calculated for a single separate custom Price-indicator (eg
the "basic").
This new single seperate Bucket-indicator can than also be referenced again, eg over and over
again in other formulas.....................and will also automaticaly be daily calculated (eg like when
added to a Chart or referenced in exploration etc. or even when OLE-ed to Excel).
Example-B.
To use Craig's example, the sum/days of U1 for an EMA U1 x-days is 1 indicator and the sum/days
of U1 for an SMAU1 x-days, is also 1 indicator and, to top it off, the U1 itselves is 1 indicator as well.
So, only separately keep the manipulated and non-manipulated values into 2x (or more) custom
indicators.
Eg, you can add yesterday's EMAU1 to today's SMAU1, then divided this outcome by 2 will give
you YOURMAU1.
Eg, creating a few more extra -now on the side- indicators should solve the problem(s).
Then playing with the above "formula idea"-structure for some kind of a guidence in building any
other formulas, should provide for an usable and working alternative for the
<onces hooked always hooked> variable (eg the un-variable) that is standard shipped (for function)
with MS.
Regards,
Ton Maas
ms-irb@xxxxxxxxxxxxx
Dismiss the ".nospam" bit (including the dot) when replying.
----- Original Message -----
From: Craig DeHaan <cdehaan@xxxxxxxxxxxx>
To: <metastock@xxxxxxxxxxxxx>
Sent: woensdag 30 juni 1999 2:05
Subject: Re: Variables
> Nico,
> I was 50% w/ you in my initial reaction. Tho Ref(U1,-1) would easily
> seem to provide previous values, U1 + Ref(U1,-1) won't do the job here.
> If Guy uses a complex variable treated (for instance) similar to an EMA
> or VMA calculation or some other nonlinear cumulative function, how
> would you propose to accommodate dropping the earlier weighted data
> effects; using Ref(U1,-1) and then adding U1 for today? It won't work
> that way for many custom formulae. You need a "bucket", as he says, to
> statically hold the value of an independent variable from some (whatever
> pertinent) previous period used in its formulation.
>
> I see Dr. Bob waiting in the wings. He'll likely code around it in a
> snap once he trains his sights on the specific target. Then we have
> Guy's technically vague momentum derivations that might exclusively
> require such a direct bedside intervention/examination. Remember, he's
> known as the doc for no other reason than this.
> Craig
>
>
> Nicholas Kormanik wrote:
> >
> > So what's wrong with using
> >
> > ref(U1,-1)
> >
> > and U1 + ref(U1,-1)
> >
> > What am I missing?
> >
> > Nicholas
>
|