Re: an old bug on the 85
[Prev][Next][Index][Thread]
Re: an old bug on the 85
stephen soltesz wrote:
>
> Hello,
>
> I recall less than a year ago someone reporting a bug on the 85
> involving complex numbers and (this is what I've forgotten) either the
> dot() or cross() product. I played with both using complex numbers and
> didn't see any problem.. but need to know what it was or what caused it,
> so as not to mess any calculations up..
>
> So, if anyone knows to what I am referring, or if it was a different
> function, let me know. I remember it relating to EE majors, for some
> reason.
>
> thankyou.
> stephen.
>
As an interesting follow-up to this bug is another manifestation I
discovered while trying to find an easy workaround.
I wanted to make two 1x2 matrixes take the first, then transpose the
second and multiply the two.. 1x2 * 2x1 and get a 1x1 answer.
As it turns out, if you transpose a complex matrix, say [[(a,b),(c,d)]]
you will get
[[(a,-b)]
[(c,-d)]]
This occurs for any dimension complex matrix MxN.
Based on this, I'd guess that the vector functions are a subset of the
matrix operations already in ROM. More specifically the transpose
function. When someone calls the dot() function the ti pretends you
have two 1xN matrixes, transposes the second and multiplies the two.
This is consistent with what I've observed, but I may be confusing the
correlation with simple causality.
I now have two questions as a result of this. I found four FAQ's on the
internet.
1. maintained by ticalc.org 2. By TI itself 3. by Ray Kremer 4. by
Stephan Bird.
answering general purpose questions about ti-8x's, yet this information
was on none of them... Why?
The second question is, does this problem manifest itself in the current
ti-86 rom.. and if so WHY!!??
Please reguard these questions as merely rhetorical.. The fact that I
have a buggy calc has gotten me a little pissed.
--
"I will take my hands and weave them to a little house,
and there you shall keep a dream--"
--Miriamne. "Winterset"
Follow-Ups:
References: