|
PureBytes Links
Trading Reference Links
|
>thanks for Your help; I haven't EMA function and so it doesn't work to me.
>Could You post EMA function please?
Sorry, I thought that was included in the TS package. I guess it's
something I added. Here it is:
{Exponential Moving Average (EMA)}
Inputs: lastvalue(NumericSimple), newdata(NumericSimple), length(NumericSimple);
Variables: w(0);
w = (2/(length+1));
EMA = newdata*w + lastvalue*(1-w);
That's it. The EMA approximates the Avg() function but is faster to
calculate, and for the purposes of ProGo and other indicators, the
results from using it instead of Avg() are virtually indistinguishable.
>----- Original Message -----
>From: "Alex Matulich" <alex@xxxxxxxxxxxxxx>
>To: "blackcloudover BillCruz" <trad_delist_anydaynow@xxxxxxxxx>
>Cc: <omega-list@xxxxxxxxxx>
>Sent: Sunday, January 20, 2002 5:13 AM
>Subject: Re: coding of Williams PRO-GO Indicator ?
>
>
>> I can't believe I have come far enough with TS to help out someone else
>> with EasyLanguage.
>>
>> Shadow wrote:
>> >My feeble attempt to EL code this simple indicator (at
>> >the bottom) is not too successful - - - any
>> >suggestions, guys ?
>>
>> Yes. You can't use an expression in a function that requires a time
>> series variable as an argument.
>> >
>> >Inputs: Len1(14)
>> >value1=Close of yesterday
>> >value2=Open of today
>> >value3=Close of Today
>> >
>> >plot1 (Avg (value2-value1,14) "Public");
>> >plot2 (value3-value2, Len1, "Pro");
>>
>> The Avg function needs a time series, not an expression that has no
>> history.
>>
>> Try this instead (typed into the e-mail without testing):
>>
>> Inputs: Len1(14);
>> Variables: pro(0), pub(0);
>> if currentbar > 1 then begin
>> pro = EMA(pro[1], Open - Close[1], Len1);
>> pub = EMA(pub[1], Close - Open, Len1);
>> end
>> else begin
>> pro = 0;
>> pub = 0;
>> end;
>> plot1(pub, "Public");
>> plot2(pro, "Pro");
>>
>> I wish you luck with this. I wasted a week investigating ProGo and
>> a lot of variants I invented. It does indicate some turns, however
>> it seemed to have more false signals than positive ones. At best
>> it's a set-up, not a signal itself.
--
,|___ Alex Matulich -- alex@xxxxxxxxxxxxxx
// +__> Director of Research and Development
// \
//___) Unicorn Research Corporation -- http://unicorn.us.com
|