|
PureBytes Links
Trading Reference Links
|
<FONT face=Arial
size=2>Robin:
<FONT face=Arial
size=2>
You're English is
fine, it's certainly better than my Italian, which is zero.
<FONT face=Arial
size=2>
I think that I
understand what you are trying to say. The problem may be is that you
don't understand the relationship between mathematics and curve fitting
(sometimes called statistics).
<FONT face=Arial
size=2>
At the web
sites, the examples are so set up that the the parabola is
predefined and predetermined. Two of the points are at the end of the latus
rectum, and the third point is at the vertex. If you take 3 points so that the
the perpendicular bisector of the line joining the first and third points does
not intersect the middle point. then these 3 points wont define a parabola,
they will define a circle.
<FONT face=Arial
size=2>
The figures at the
web site predefines the parabola. Referring to the figure at the 3rd web
site. in general
FP1 <>
P1D1
<FONT face=Arial
size=2>
As I wrote earlier,
data points do not lie on nice smooth curves. When curve fitting data, we have
to decide what kind of curve fits the data and then try to fit such a curve to
the data. For these reasons we need many more points than the minimum
needed to define the curve.
<FONT face=Arial
size=2>
If you have access
to an academic library look up some books on curve fitting. Try to find
some that were written before the days of computers. Let me know what you
find.
<FONT face=Arial
size=2>
<FONT face=Arial
size=2>Lionel
<FONT face=Arial
size=2>
<FONT face=Arial
size=2>
<FONT face=Tahoma
size=2>-----Original Message-----From: Robin Hood
[mailto:robinhood@xxxxxxxx] Sent: Sunday, January 05, 2003 9:51
AMTo: Metastockusers@xxxxxxxxxxxxxxxSubject: Re:
[Metastockusers] RE:Re: Bar Count since certain date
Lionel,
my bad English doesn't help us... :-)
Look at rhis site where there are a full description of what you already
know, I think
<A
href="http://schools.spsd.sk.ca/mount/Hoffman/MathC30/Parabola/parabola.htm#trinomial";>http://schools.spsd.sk.ca/mount/Hoffman/MathC30/Parabola/parabola.htm#trinomial
I find this about autocad (I don't use it)
<A
href="http://xarch.tu-graz.ac.at/autocad/news/lisp_progs/msg00013.html";>http://xarch.tu-graz.ac.at/autocad/news/lisp_progs/msg00013.html
and anyway the results of this search say that wee need three points for
building a parabola (by solving a system of equations I remember - maybe :-)
)
<A
href="http://www.google.it/search?q=build+parabola+three+points&hl=it&lr=&ie=UTF-8&oe=UTF-8&start=0&sa=N";>http://www.google.it/search?q=build+parabola+three+points&hl=it&lr=&ie=UTF-8&oe=UTF-8&start=0&sa=N
I hope your math knowledge is better than mine and unserstand what I'd like
to say.
robin
<BLOCKQUOTE
style="PADDING-RIGHT: 0px; PADDING-LEFT: 5px; MARGIN-LEFT: 5px; BORDER-LEFT: #000000 2px solid; MARGIN-RIGHT: 0px">
----- Original Message -----
<DIV
style="BACKGROUND: #e4e4e4; FONT: 10pt arial; font-color: black">From:
Lionel
Issen
To: <A
title=Metastockusers@xxxxxxxxxxxxxxx
href="mailto:Metastockusers@xxxxxxxxxxxxxxx";>Metastockusers@xxxxxxxxxxxxxxx
Sent: Sunday, January 05, 2003 3:13
PM
Subject: RE: [Metastockusers] RE:Re: Bar
Count since certain date
<FONT face=Arial
size=2>Robin:
<FONT face=Arial
size=2>
Can you direct me
to the portion of the cited web site where it says that 3 points are
sufficient to define a parabola? Three points would be sufficient to define a
parabola if you know the semi latus rectum (factor F
below).
<FONT face=Arial
size=2>
Please bear with
me. While both the circle and the parabola are conic sections their equations
are different.
<FONT face=Arial
size=2>
Equation of Circle
: X**2 + Y**2 = R**2 R is the radius
<FONT face=Arial
size=2>
Equation of
parabola: Y**2 =
4FX F is the
semi latus rectum
<FONT face=Arial
size=2>
Here is a little
exercise you can do. All you need are two pieces of paper, a
pencil, a compass, a straight edge (a ruler will do) and maybe an
eraser.
Place 3 points on
a piece of paper, they must not be on a straight line. Call them A, B, and C.
Draw the lines A-B and B-C. Bisect these 2 lines and draw the perpendicular
bisectors. the bisectors will meet at the center of the circle. Next select
any other 3 points on the circle and repeat this little
exercise. These bisectors will also meet at the center of the
circle. Before the days of electronic navigation, the basis of
this method was used by navigators near shore or in narrow
channels to locate the position of the vessel.
<FONT face=Arial
size=2>
If you draw a
parabola and select two or more sets of three points on the parabola
and repeat the exercise, the bisectors will not meet at the same
place.
<FONT face=Arial
size=2>
If you have any
further questions please write me privately
<FONT face=Arial
size=2>
<FONT face=Arial
size=2>Lionel
<FONT face=Arial
size=2>
<FONT face=Arial
size=2>
<FONT face=Arial
size=2>
<FONT
face=Tahoma size=2>-----Original Message-----From: Robin Hood
[mailto:robinhood@xxxxxxxx] Sent: Sunday, January 05, 2003 4:51
AMTo: Metastockusers@xxxxxxxxxxxxxxxSubject: Re:
[Metastockusers] RE:Re: Bar Count since certain date
Hi Lionel
Three points are sufficient as parabola is made by tho points
equidistanced (? in english ?) form another one.
<A
href="http://www.xahlee.org/SpecialPlaneCurves_dir/Parabola_dir/parabola.html";>http://www.xahlee.org/SpecialPlaneCurves_dir/Parabola_dir/parabola.html for
examples (by google).
Here is a chart form an Italian site which I don't name because of
netiquette (I think...). They build a parabola from three points, in effect.
It's the same effect I'd like to reproduce with Metastock... but the trouble
are the input dates (x coordinates) :-(
robin
<BLOCKQUOTE
style="PADDING-RIGHT: 0px; PADDING-LEFT: 5px; MARGIN-LEFT: 5px; BORDER-LEFT: #000000 2px solid; MARGIN-RIGHT: 0px">
----- Original Message -----
<DIV
style="BACKGROUND: #e4e4e4; FONT: 10pt arial; font-color: black">From:
Lionel
Issen
To: <A
title=Metastockusers@xxxxxxxxxxxxxxx
href="mailto:Metastockusers@xxxxxxxxxxxxxxx";>Metastockusers@xxxxxxxxxxxxxxx
Sent: Saturday, January 04, 2003 4:57
PM
Subject: RE: [Metastockusers] RE:Re:
Bar Count since certain date
Three points
always define a circle. I think that you need at least 4 points to
define a parabola. When curve fitting a parabola you need more than 4 points
as data points always have some noise, they don't fall on a nice smooth
curve.
<FONT
face=Tahoma size=2>-----Original Message-----From: Robin Hood
[mailto:robinhood@xxxxxxxx] Sent: Saturday, January 04, 2003 8:25
AMTo: Metastockusers@xxxxxxxxxxxxxxxSubject: Re:
[Metastockusers] RE:Re: Bar Count since certain date
Hi Group
Thanks to Spyros for his formula.
In order to build a parabolic trend-line, I need three points (geometry
requires it). It means three x,y (date, price) coordinates.
The problem is that metastock accepts only six inputs (in oder to
define dates I need 3+3+3=9 inputs :-( ).
Is there another way to make this?
Now, the formula inputs are as follows, but I'd like to change the
numbers with dates since I don't understand very well and immediately their
sequence
x1:=Input("point 1 : x1",-100000000,100000000,0);y1:=Input("
y1",-100000000,100000000,0);x2:=Input("point 2 :
x2",-100000000,100000000,0);y2:=Input("
y2",-100000000,100000000,0);x3:=Input("Point 3:
x3",-100000000,100000000,0);y3:=Input("
y3",-100000000,100000000,0);
Thanks in advance, Robin
<BLOCKQUOTE
style="PADDING-RIGHT: 0px; PADDING-LEFT: 5px; MARGIN-LEFT: 5px; BORDER-LEFT: #000000 2px solid; MARGIN-RIGHT: 0px">
----- Original Message -----
<DIV
style="BACKGROUND: #e4e4e4; FONT: 10pt arial; font-color: black">From:
SR
To: <A
title=Metastockusers@xxxxxxxxxxxxxxx
href="mailto:Metastockusers@xxxxxxxxxxxxxxx";>Metastockusers@xxxxxxxxxxxxxxx
Sent: Saturday, January 04, 2003 2:24
PM
Subject: [Metastockusers] RE:Re: Bar
Count since certain date
bruneski came up with a variation of my approach which
is corrects somefailures of my version.So I suggest trying this
one:{BarsSinceDate}d:=Input("Day",1,31,1);m:=Input("Month",1,12,1);y:=Input("Year",1920,2030,2002);mydate:=
10000*y + 100*m + d;date:= 10000*Year() + 100*Month() +
DayOfMonth();count:=BarsSince(mydate<=date); {use '<' if first
day is not to becounted}count:=count +
Cum(If(count=0,1,0));countTo unsubscribe from
this group, send an email
to:Metastockusers-unsubscribe@xxxxxxxxxxxYour
use of Yahoo! Groups is subject to the <A
href="http://docs.yahoo.com/info/terms/";>Yahoo! Terms of Service.
To unsubscribe from this group, send an email
to:Metastockusers-unsubscribe@xxxxxxxxxxxYour
use of Yahoo! Groups is subject to the <A
href="http://docs.yahoo.com/info/terms/";>Yahoo! Terms of Service.
To unsubscribe from this group, send an email
to:Metastockusers-unsubscribe@xxxxxxxxxxxYour
use of Yahoo! Groups is subject to the <A
href="http://docs.yahoo.com/info/terms/";>Yahoo! Terms of Service.
To unsubscribe from this group, send an email
to:Metastockusers-unsubscribe@xxxxxxxxxxxYour use
of Yahoo! Groups is subject to the <A
href="http://docs.yahoo.com/info/terms/";>Yahoo! Terms of Service.
To unsubscribe from this group, send an email
to:Metastockusers-unsubscribe@xxxxxxxxxxxYour use
of Yahoo! Groups is subject to the <A
href="http://docs.yahoo.com/info/terms/";>Yahoo! Terms of Service.
To
unsubscribe from this group, send an email
to:Metastockusers-unsubscribe@xxxxxxxxxxxYour use
of Yahoo! Groups is subject to the <A
href="http://docs.yahoo.com/info/terms/";>Yahoo! Terms of Service.
Yahoo! Groups Sponsor
ADVERTISEMENT
<img
src="http://us.a1.yimg.com/us.yimg.com/a/am/ameritrade/120402_am_ban_off_x82_x_300x250_6.gif"; width=300 height=250 border=0>
To unsubscribe from this group, send an email to:
Metastockusers-unsubscribe@xxxxxxxxxxx
Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.
|
![http://us.a1.yimg.com/us.yimg.com/a/am/ameritrade/120402_am_ban_off_x82_x_300x250_6.gif"](http://us.a1.yimg.com/us.yimg.com/a/am/ameritrade/120402_am_ban_off_x82_x_300x250_6.gif"){kind=link}