|
PureBytes Links
Trading Reference Links
|
Paul,
Let's see if we can provide some recap information for anyone else
wishing to take part in this. The indicators were listed by you in
message 24953.
As I remember we had several indicators from which to take signals. We
had a MACD that we made using the ZeroLag EMA and we also took the
ZeroLag MACD and normalized it.
Now we want to do an exploration and optimize some values to see what
performs best.
This was always a lot of fun but also consumed large amounts of time.
Even so you have to walk through it just to see how it works.
Optimization is not widely favored, so just realize that there are a
lot of critics out there. There are no hardened rules to optimizing as
any value is fair game but just realize that it is far better to run
your test on a smaller number of values and if you have to run your
test several times thats quite alright.
You have chosen the ZeroLag EMA and a moving average crossover. Easy
enough. First go into the system tester and first thing you notice is
that there already are some test included. Let's take a look at the
Equis-moving average crossover...open it. Open the Buy Order tag. What
you will see is:
Mov(C,opt1,E) > Mov(C,opt2,E)
Next click on the Optimizations tag and look at the values assigned
for opt1 and opt2.
In order to use the Zero Lag EMA all you will need to do is replace
the Close or C with an assigned variable for the fml("Zero Lag EMA").
Like this:
A:= fml("Zero Lag EMA");
Now just reference it in your system test. Like this:
A:= fml("Zero Lag EMA");
A > Mov(A,opt1,E)
This will leave the original formula intact and you will be optimizing
on the moving average of it. If you want to optimize the original
formula you must include it in your test formula instead of just the
fml call variable. Notice in the example above that all we did was
place an opt# where a numeric value would normally go.
I would be careful about over-optimizing though. The results are
really going to be the result of how well the formula performs on a
particular stock or group of stocks during a particular period of time
and may not be an indication of how well future performance can be
determined. Best to try the test on stocks of varying performance and
use those as a benchmark. This is really where most criticism comes
from.
That should get you going, Let us know how it goes.
Preston
--- In equismetastock@xxxxxxxxxxxxxxx, "Paul Harris" <paul_vicmar@xxx>
wrote:
>
> Some weeks ago Preston and I worked on an exploration and expert
> advisor for a Zero Lag MACD.
> So I was looking at using a Zero Lag EMA in a system test, similar to
> a MA crossover. My only problem is that I want to use the optimiser
to
> find the best time periods for Zero Lag EMA and I don´t know how to
do it.
> I know I have to indentify the Zero Lag EMA as fml("Zero Lag EMA")but
> then how can I introduce the opt1 function?
> Some help would be greatly appreciated.
> Thanxs
> PAUL
>
Yahoo! Groups Links
<*> To visit your group on the web, go to:
http://groups.yahoo.com/group/equismetastock/
<*> Your email settings:
Individual Email | Traditional
<*> To change settings online go to:
http://groups.yahoo.com/group/equismetastock/join
(Yahoo! ID required)
<*> To change settings via email:
mailto:equismetastock-digest@xxxxxxxxxxxxxxx
mailto:equismetastock-fullfeatured@xxxxxxxxxxxxxxx
<*> To unsubscribe from this group, send an email to:
equismetastock-unsubscribe@xxxxxxxxxxxxxxx
<*> Your use of Yahoo! Groups is subject to:
http://docs.yahoo.com/info/terms/
|