> Graphing doesn't like a two-dimensional value, such as a complex number,
>since it can't plot that much information with a single pixel. Try plotting
>a real-valued version of the function, such as RE(X^(1/3)) or ABS(X^(1/3))
>and see if that helps show what you want to see.
Fot course this will not give the wanted results, because RE(-5^(1/3))
<> XROOT(-5,3). (Err, or am I goofing up now?)
JPSOFT
--
Jarno Peschier, jpes...@cs.ruu.nl, 125:3101/100.1@WEM,
57:3105/205@IntlNet, 88:950/1.8@BasicNet.
___________________________________________________________________________
Life would be so easy, if only we had the sourcecode...
: -----------------------------------------------------------------------
: Mark Sztainbok
: (mjs...@lindblat.cc.monash.edu.au)
Oh, it does plot them, but it does so in the imaginary plane. That's a
plotting surface perpendicular to the display on your calculator. The
pixels and wiring necessary to show such a plane are rather difficult
to install, and they break off easily. So HP omitted them.
I would imagine that, if enough users complained long and loudly about
this shortcoming, HP would design a fix into the next ROM revision, and
we could all send in our 48s for free upgrades.
Regards
Ray Depew
r...@fc.hp.com
Hint: If -1 = i^2, then (-1)^(1/3) = i^(2/3). Where/how would you
plot that in an X-Y coordinate system?
>Recently I was helping someone with some graph work and found that if I
>plotted a graph such as x^(1/3) or x^(1/5), only half of the graph
>(x>=0) would appear. Is there any way to get the whole graph to appear
>without using XROOT (So that functions such as F' and AREA will still
>work)? Why does only half the graph appear?
You really found "something strange" here. The reason that x^(1/3)
does not get drawn <0 is because the ^ then gives complex results (not
then solution we usually want, but the complex number with the
smallest angle). Try it by entering e.g. '5^(1/3)' EVAL on the stack.
Complex numbers are not draw on a normal X-Y plot.
Maybe you can convince the ^ function to give back real results to
x^(1/n) with x<0 with some flag, but I can't try that now, because I
get an "Error: Insufficient Memory" when trying to acces the flag-list
(I only have a G so far and I like an occasional game...).
It works but it takes a few minutes to set up. That is why i don't use
it any more. That and you have to determine when the function is
imaginary.
I am a mechanical engineering student so this is probably not the easiest
way around the problem.
ARS
Dave.
------------------------
All disclaimers apply. This ain't an official anything. I only work here.
You x^(1/3) is only defined for x<0 if (1/3) is EXACTLY 1/3. For
example, (-8)^0.3333333 =(-8)^(3333333/10000000), which has an odd
numerator, and the answer would be close to -2. (-8)^0.3333334, on the
other hand would be close to 2, as it is equal to the 1000000th root of
-8 to an even power, with is positive.
--
| Darwin Hawes <ha...@physics.ubc.ca>
|
| Computers are called stupid because they follow instructions to the letter.
| People are called stupid if they don't.