|
PureBytes Links
Trading Reference Links
|
Dans un courrier daté du 20/09/98 01:17:03 , vous avez écrit :
<<
It's a pretty sad comment when evidence shows a competitor going to
greater lengths to address a Trade Station user's concerns than does
Omega.
No wonder we're seeing such positive reaction (even from Pierre :-)
toward banding together to advertise the existance of the Omega List to
newbies. What better way than that to get Omega to understand Customers
are not Mushrooms...to be kept in the Dark and fed BS as if they were
some 4H Project run amuck.
Dave
> Subject:
> RE: TL_VB & EL Square Root Functions
> Date:
> Fri, 18 Sep 1998 21:19:31 -0400
> From:
> "Mark J. Cerar" <mcerar@xxxxxxxxx>
> To:
> "Scientific Approaches" <sci@xxxxxxxxxx>, "TradeLab Mail List"
<TradeLab@xxxxxxxxxx>
>
>
> Early this year, after having read a piece in "Technical Analysis of Stocks
> & Commodities" ( p. 67, January, 1998), I tried to reproduce the
Correlation
> function in TS. In order to calculate the correlation coefficient you need
> to calculate the square root of the variance. I could not get the results
> of my TS function to match the results obtained in Excel using identical
raw
> data. When I wrote to EasyLang Support, the technical support rep simply
> brushed me off saying there was a problem in my code and that they don't
> correct users faulty program code. He told me that one of the Omega
> Solution Providers could help me. Now I understand why I couldn't work but
> couldn't the rep simply have told me this instead of feeding me BS.
>
> Mark.
Well...
I have had a look at Jan98 issue of TASC (p 67), and saw nothing wrong that
could not be coded in TS.
Variance and Covariance are calculation similar to Standard deviation (when
considereing the maths involved) , and the correlation coefficient is very
simple to calculate, the squareroot function being involved once in
(V(x)*V(y))^1/2.
The discussion ater was dealing with single precision vs double when
comparing EL and VB for the squareroot function and error trapping, and I
seriously doubt that the EL error came from here.
As we do not have the Excel sheet and the EL code from this customer, it's
impossible to draw the kind of conclusion that you are accustomed to serve us
without any proof.
For your information, variance and covariance function have been added to the
TS5 power editor, as well as the alternate stddev function discussed here one
year ago.
Rgds,
Pierre Orphelin
|