home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!dtix!darwin.sura.net!wupost!gumby!destroyer!news.iastate.edu!vincent2.iastate.edu!brianc
- From: brianc@iastate.edu (Brian J Cerveny)
- Newsgroups: comp.sys.hp48
- Subject: Re: Hp48 and TI85 compared
- Message-ID: <brianc.716174465@vincent2.iastate.edu>
- Date: 11 Sep 92 01:21:05 GMT
- References: <1992Aug14.210022.15328@utagraph.uta.edu> <1992Aug20.023308.6024@unixg.ubc.ca> <1992Aug20.215637.9692@utagraph.uta.edu>
- Sender: news@news.iastate.edu (USENET News System)
- Organization: Iowa State University, Ames IA
- Lines: 97
-
- In <1992Aug20.215637.9692@utagraph.uta.edu> b645zjo@utarlg.uta.edu (Rick Homard) writes:
-
- >>In article <1992Aug20.023308.6024@unixg.ubc.ca>, ochealth@unixg.ubc.ca (ochealth) writes...
-
- >>>Unit Management
- >>>
- >>>Not supported on the TI85.
- >>
- >>I assume this means you can't define your own? Can you perform arithmetic
- >>with units?
- >>>
- >TI doesn't call them units. Rather it is just simple conversion between
- >units.
- >And yes you can define your own unit conversions.
-
- Then what do you mean by Unit Management?
-
- >I compared the TI85's root finder with the HP48 (using the root finding
- >program supplied by Bill Wickes on the net). The results are incredible.
- >The same 30th order polynomial was input to each machine. The HP48 found
- >all of the roots in 80 seconds. I stopped the TI85 after 15 minutes.
- >It is obvious that the TI85 gets heavily labored when required to perform
- >at its limits.
-
-
- >>>The TI85 performs some fudging (rounding, et al) on calculations.
- >>>
- >>>Examples:
- >>>
- >>>0.333333333333 * 3 = 1
-
- Funny, with Float set in the MODE menu, I get .999999999999. If you round to
- any finite number, you will get 1.(n-1), where n=fixed point. Make sure you
- have FLOAT set.
-
- >>>and
- >>>sin (3.14159265359) = 0
- >
- Same here, I get -2.000000000000E-13 when I enter that in my machine.
-
- >>If I remember TI, they don't really round the answer, they just display
- >>part of the answer (ie the first 12 digits, and the 13 and 14 are kept
- >>internally). Sort of like the HP with 10 FIX on all the time, but trailing
-
- As mentioned above, it's not really like 10 FIX (the equivalent on TI would be
- 11 fixed) unless there are more than 12 digits in the result, AND the
- thirteenth digit is >=5 after any internal rounding is done.
- For example:
-
- .44444444444444 displays as-is.
- .44444444444445 displays as-is (since the 5 is in the fourteenth place,
- not affecting the 12th)
- .44444444444454 displays as .444444444445. The fourteenth digit is OK,
- but the five in the 13th place rounds the 12th displayed
- digit.
-
-
- >This is still "fudging" notheless. Regardless of how many decimal places
- >you take .33333333 , when you multiply it by 3 you get .999999999 to the
- >same number of decimal places.
- >Also sin(3.14159265359) is a small number but not zero. The sin(pi)
- >is zero but 3.14159265359 is not pi.
-
- Right, and my TI-85 agrees. Check your setup.
-
- >>How fast is it? Does it return an error estimate? I had a TI56 that did
- >>Simpson's rule, and it was SLOW. When it finished you had NO CLUE whether
- >>the answer is meaningful. The HP uses an adaptive algorithm which will
- >>usually return usefull results to an integral, and give you a good idea
- >>whether the numerical result is close, or way off.
-
- >It is fast and it returns an error estimate in a variable, just like
- >the HP48. As what method it uses I do not know.
-
- My TI-68 uses Simpson's Rule, and since the 68 is much like the 85 in non-
- graphical or programming function, the TI-85 probably does to.
-
- >>Is there any way to "get inside" the TI85? I mean software wise. A lot
- >>of power in the HP48 comes from being able to add low level machine language
- >>programs that blow away any user level program. For example, the recent ML
- >>FFT is blindingly fast, and it can't be done quickly in user code.
- >>
- >I do not know. I have only had it for about a week now.
-
- I'm curious about this too. There is no mention whatsoever about this in the
- manual, so I doubt it. But you never know. :)
-
- I tried calling the technical help line but they were closed. (The tech line
- is open from 8a - 4:20p, and is a long distance call... :( What's the HP
- tech support like?) I wanted to ask about low-level programming, and the
- absence of I/O port programmability. Since port I/O isn't a part of the
- high-level language, it would have to be available from machine-level,
- which is at this time a questionable prospect.
-
- --
- // Brian J. Cerveny (brianc@iastate.edu)
- \X/ Disclaimer: The views expressed above are my own. So there.
-