|
PureBytes Links
Trading Reference Links
|
Hi Alex,
The comment was a joke, not a jibe. Hope you took it that way. A
couple of people suggested commercial indicators/functions that may or
may not have been suitable. I was simply wondering if there was a
formula for such things, bit like the linear reg stuff. Thank you very
much for your comments and formula.
You - and a lot of the guys on the list - have been extremely helpful
many times in many ways and it's very much appreciated.
Ian
> >Thanks to everyone who replied. A few suggested smoothing functions
> but >that's not really what I wanted to do. Anyway, I've been given
> to >understand that it can be done although possibly not in TS but
> alas, no one could tell me how to do it - at least not for free...
> (:-)
>
> I can give you some advice for free.
>
> You can often make a curve fit into a linear regression. For example,
> if you have exponentially rising data that you suspect is of the form
> y=a*exp(mx+b), then you can convert it to a line u=mx+b, where
> u=ln(y). Then you find the fit for slope (m) and intercept (b), and
> substitute back.
>
> Similarly, if you suspect your data has the form y = sin(p(x+b)),
> you can invert it as u = px+pb where u=arcsin(y), and solve fo the
> slope p and intercept pb.
>
> That's my $0.02.
>
> I wouldn't do it for free either, btw.
>
> -Alex
>
>
|