k3u@mentor.cc.purdue.edu (Barry Smith) (12/03/90)
Does anyone have the source code to routines that can extract the complex roots of polynomials? If there is a book that has the source code that would be fine too. -Barry Smith
ts@uwasa.fi (Timo Salmi) (12/04/90)
In article <2173@mentor.cc.purdue.edu> k3u@mentor.cc.purdue.edu (Barry Smith) writes: >Does anyone have the source code to routines that can extract >the complex roots of polynomials? If there is a book that has >the source code that would be fine too. If you want the source code you can find it in TP numerical toolbox. If a program without a source code is sufficient try getting /pc/ts/tsnum12.arc by anonymous ftp from uwasa.fi. The package contains among others a polyr.exe program for the said purpose. ................................................................... Prof. Timo Salmi (Moderating at anon. ftp site 128.214.12.3) School of Business Studies, University of Vaasa, SF-65101, Finland Internet: ts@chyde.uwasa.fi Funet: gado::salmi Bitnet: salmi@finfun
hagins@gamecock.rtp.dg.com (Jody Hagins) (12/06/90)
In article <2173@mentor.cc.purdue.edu>, k3u@mentor.cc.purdue.edu (Barry Smith) writes: |> Does anyone have the source code to routines that can extract |> the complex roots of polynomials? If there is a book that has |> the source code that would be fine too. |> |> -Barry Smith |> Yeh, I wrote one when I was in school too. I didn't have any problems with it, and as I remember, even the "slower" students seemed to provide a workable solution. So, keep working (or should I say start working), and I'm sure you'll be able to write one soon enough. GENERAL NOTE: It's getting close to finals time, and while it is tempting, this newsgroup is not the place for getting your homework done. This group (comp.lang.c) and comp.lang.pascal (the other group to which this article was posted) is meant for mutual exchange (and education) concerning these languages. Not only is it unethical, but I'm sure your school has an honor code which prohibits this type of activity. I'm getting tired of seeing "do my homework" requests on this network!!!! How in the hell do you expect to learn anything if others do it for you all the time?!?!?!?!? -- Jody Hagins hagins@gamecock.rtp.dg.com Data General Corp. 62 Alexander Dr. RTP, N.C. 27709 (919) 248-6035
subbarao@phoenix.Princeton.EDU (Kartik Subbarao) (12/06/90)
In article <1990Dec5.175558.29859@dg-rtp.dg.com> hagins@gamecock.rtp.dg.com (Jody Hagins) writes: >In article <2173@mentor.cc.purdue.edu>, k3u@mentor.cc.purdue.edu (Barry Smith) writes: >|> Does anyone have the source code to routines that can extract >|> the complex roots of polynomials? If there is a book that has >|> the source code that would be fine too. > >Yeh, I wrote one when I was in school too. >I didn't have any problems with it, and as I remember, even >the "slower" students seemed to provide a workable solution. >So, keep working (or should I say start working), and I'm >sure you'll be able to write one soon enough. > > >GENERAL NOTE: > >It's getting close to finals time, and while it is tempting, >this newsgroup is not the place for getting your homework >done. This group (comp.lang.c) and comp.lang.pascal (the other >group to which this article was posted) is meant for mutual exchange >(and education) concerning these languages. Not only is it unethical, >but I'm sure your school has an honor code which prohibits this >type of activity. I'm getting tired of seeing "do my homework" >requests on this network!!!! How in the hell do you expect to >learn anything if others do it for you all the time?!?!?!?!? Yes this is true. But why are you assuming that he wants to do his homework on USENET? Don't you think its a bit prejudging to assume right off the bat about a person's intentions? Especially the part about "slower" students and the like. That's getting downright condescending. I really don't think that kind of tone belongs in an educational newsgroup (your own words). What if he is doing some independent work on the stuff and wants to find out whether there already exists code to do what he wants? There certainly is no point to re-inventing the wheel. I certainly am not going to take too much time writing someting that I need done quickly and is already around. IMNSFHO comp.lang.c is an informational newsgroup about C, and people are here for free exchange of information regarding the language. There is no need for censors like you to say "oh that question shouldn't be answered", because it comes from an undergraduate student. This might make you feel sad, but believe it or not, I have learned stuff on this newsgroup that has helped me with (gosh) official computer programming assignments that I hand in. I think that's the purpose of this, not to decide who to dole out information to. If a person wants to do his homework on the net, there really is no way to stop him. It's between him and his own conscience as to what he wants, and others can't make that decision for him. If you don't want to answer him, then don't. But there is no need to collectively put down questions of this sort under a WARNING: flag. Anyway -- what does everyone else think about this? I've seen both sorts of attitudes on the subject. What do you guys think comp.lang.c should do? -Kartik (I need a new .signature -- any suggestions?) subbarao@{phoenix or gauguin}.Princeton.EDU -|Internet kartik@silvertone.Princeton.EDU (NeXT mail) -| SUBBARAO@PUCC.BITNET - Bitnet
ron@argus.lpl.arizona.edu (Ron Watkins) (12/07/90)
I have seen some argument on this issue before in this group. There are surely good reasons for which to NOT do someone elses homework. However, its difficult to determine when such a condition arises. For example: I work at the U. of Ariz. and have had the need to use hash functions. Just because I inquire the net about hash function and performance AND I have a university mail address AND its final's time doesn't mean it's homework! I agree with previous posts that if you feel that someone is doing their homework on the net, then don't answer. No need to bash heads :-> Ron Watkins ron@argus.lpl.arizona.edu
fritz@urz.unibas.ch (12/07/90)
In article <2173@mentor.cc.purdue.edu>, k3u@mentor.cc.purdue.edu (Barry Smith) writes: > Does anyone have the source code to routines that can extract > the complex roots of polynomials? If there is a book that has > the source code that would be fine too. > > -Barry Smith For a lot of numerical problems I can STRONGLY recommend the book NUMERICAL RECIPES IN PASCAL -- The Art of Scientific Computing written by four authors (Press, Flannery et.al.) and published by Cambridge University Press (UK). It contains many procedure listings and you can order an 51/4" disk with all the nice stuff on it. The book is very good from the mathematical point of view and it is fun to read! Oliver Fritz, University of Basel, Switzerland PS: If a student wanted to have his homework done by someone else he wouldn't have to post it on the net. Probably his professor would read it there quite soon...
rh@smds.UUCP (Richard Harter) (12/07/90)
In article <4529@idunno.Princeton.EDU>, subbarao@phoenix.Princeton.EDU (Kartik Subbarao) writes: > In article <1990Dec5.175558.29859@dg-rtp.dg.com> hagins@gamecock.rtp.dg.com (Jody Hagins) writes: > >In article <2173@mentor.cc.purdue.edu>, k3u@mentor.cc.purdue.edu (Barry Smith) writes: > >|> Does anyone have the source code to routines that can extract > >|> the complex roots of polynomials? If there is a book that has > >|> the source code that would be fine too. [Snide comments which make unwarranted assumptions about the request being one for meeting homework deleted.] > Yes this is true. But why are you assuming that he wants to do his homework > on USENET? Don't you think its a bit prejudging to assume right off the > bat about a person's intentions? Especially the part about "slower" > students and the like. That's getting downright condescending. I really > don't think that kind of tone belongs in an educational newsgroup (your own > words). What if he is doing some independent work on the stuff and wants to > find out whether there already exists code to do what he wants? There > certainly is no point to re-inventing the wheel. I certainly am not going > to take too much time writing someting that I need done quickly and is > already around. Perhaps it is childish of me, but I was highly offended by Jody's comments, which I felt were snide, unwarranted, and, considering the specific request, silly. First of all many people who are not computer science students taking a numerical analysis course have reason to need a routine to find the complex roots of a polynomial -- for example, physicists, electrical engineers, chemists, mechanical engineers, etc. It's a very common problem in applied math in all fields. Secondly it's a hard problem. Quite frankly I doubt that Jody has ever written a good general purpose routine for finding complex roots of polynomials. I would be surprised that anyone who has had to deal seriously with the problem would dismiss it so casually and comptemptuously. Most of the methods given in the numerical analysis texts fail badly on ill-conditioned polynomials. In response to the original query -- your best option, if it is available, is to cast the problem as an eigenvalue problem and use a canned eigenvalue package to get the roots. Your system should have one available. If not I believe you can get the EISPACK package via anonymous ftp. In any case the best groups for this request are comp.lang.fortran, comp.math, and comp.math.num-analysis (sp?). Also check to see whether the IMSL routines are available on your system. -- Richard Harter, Software Maintenance and Development Systems, Inc. Net address: jjmhome!smds!rh Phone: 508-369-7398 US Mail: SMDS Inc., PO Box 555, Concord MA 01742 This sentence no verb. This sentence short. This signature done.
dmurdoch@watserv1.waterloo.edu (D.J. Murdoch - Statistics) (12/07/90)
In article <1990Dec6.192941.1215@urz.unibas.ch> fritz@urz.unibas.ch writes: > >For a lot of numerical problems I can STRONGLY recommend the book > > NUMERICAL RECIPES IN PASCAL -- The Art of Scientific Computing > >written by four authors (Press, Flannery et.al.) and published by >Cambridge University Press (UK). I have the original Numerical Recipes, which has the code in Fortran with fairly literal Pascal translations in an appendix. Have you (or anyone else) looked at both versions enough to be able to recommend for or against buying the pure Pascal version? Duncan Murdoch
eric@wdl47.wdl.fac.com (Eric Kuhnen) (12/08/90)
subbarao@phoenix.Princeton.EDU (Kartik Subbarao) writes: >Yes this is true. But why are you assuming that he wants to do his homework >on USENET? Don't you think its a bit prejudging to assume right off the >bat about a person's intentions? Especially the part about "slower" >students and the like. That's getting downright condescending. I really >don't think that kind of tone belongs in an educational newsgroup (your own >words). What if he is doing some independent work on the stuff and wants to >find out whether there already exists code to do what he wants? There >certainly is no point to re-inventing the wheel. I certainly am not going >to take too much time writing someting that I need done quickly and is >already around. A valid point. I, too, felt as though the response was a bit condenscending. >IMNSFHO comp.lang.c is an informational newsgroup about C, and people are >here for free exchange of information regarding the language. There is no >need for censors like you to say "oh that question shouldn't be answered", >because it comes from an undergraduate student. This might make you feel >sad, but believe it or not, I have learned stuff on this newsgroup that >has helped me with (gosh) official computer programming assignments that >I hand in. I think that's the purpose of this, not to decide who to dole >out information to. >If a person wants to do his homework on the net, there really is no way to >stop him. It's between him and his own conscience as to what he wants, and >others can't make that decision for him. If you don't want to answer him, >then don't. But there is no need to collectively put down questions of this >sort under a WARNING: flag. >Anyway -- what does everyone else think about this? I've seen both sorts of >attitudes on the subject. What do you guys think comp.lang.c should do? In defense of the first reply, I might say that the original poster did ask for an answer to what appeared to be a homework assignment. I think, however, that any "I had to do it so you have to do it, too" attitude that is handled in the rather cavalier manner of the first reply is unjustified. I think a more gentle way of approaching this situation would have been to reply, "I feel uncomfortable giving away homework assignments. Is your request for source code driven because you have a homework assignment to finish? If so, I would rather not give away the answer. However, here is a general algorithm that I would be happy to answer questions on. Fair enough?" On a related note, I tend to believe that information should be exchanged freely. I don't think that the state of the art in computer programming would be higher if without the code and idea borrowing that happens in labs, hacker closets, and on this net. In fact, I would love to teach a course that builds some complex program almost entirely upon borrowed code. "Q" > >(I need a new .signature -- any suggestions?) >subbarao@{phoenix or gauguin}.Princeton.EDU -|Internet >kartik@silvertone.Princeton.EDU (NeXT mail) -| >SUBBARAO@PUCC.BITNET - Bitnet
cdm@gem-hy.Berkeley.EDU (Dale Cook) (12/08/90)
In article <4529@idunno.Princeton.EDU>, subbarao@phoenix.Princeton.EDU (Kartik Subbarao) writes: |> |> IMNSFHO comp.lang.c is an informational newsgroup about C, and people are |> here for free exchange of information regarding the language. There is no |> need for censors like you to say "oh that question shouldn't be answered", |> because it comes from an undergraduate student. This might make you feel |> sad, but believe it or not, I have learned stuff on this newsgroup that |> has helped me with (gosh) official computer programming assignments that |> I hand in. I think that's the purpose of this, not to decide who to dole |> out information to. |> I agree. |> If a person wants to do his homework on the net, there really is no way to |> stop him. It's between him and his own conscience as to what he wants, and |> others can't make that decision for him. If you don't want to answer him, |> then don't. But there is no need to collectively put down questions of this |> sort under a WARNING: flag. |> If a person wants someone to do their homework for them, so be it. They're shortchanging themselves. |> Anyway -- what does everyone else think about this? I've seen both sorts of |> attitudes on the subject. What do you guys think comp.lang.c should do? |> |> |> |> -Kartik I've never understood the "net censorship" mentality. If you feel that someone is trying to get a homework problem done on the net and you don't want to help them, DON'T ANSWER. This group should be a forum where one feels free to ask questions. When people don't feel free to ask questions, no matter how seemingly trivial to some, the purpose of the group is lost. ---------------------------------------------------------------------- --- Dale Cook cdm@inel.gov "The only stupid question is the unasked one." The following disclaimer was added by my employer. No flames, please. ---------------------------------------------------------------------- ========== long legal disclaimer follows, press n to skip =========== ^L Neither the United States Government or the Idaho National Engineering Laboratory or any of their employees, makes any warranty, whatsoever, implied, or assumes any legal liability or responsibility regarding any information, disclosed, or represents that its use would not infringe privately owned rights. No specific reference constitutes or implies endorsement, recommendation, or favoring by the United States Government or the Idaho National Engineering Laboratory. The views and opinions expressed herein do not necessarily reflect those of the United States Government or the Idaho National Engineering Laboratory, and shall not be used for advertising or product endorsement purposes.
bigelow@hpfcso.HP.COM (Jim Bigelow) (12/08/90)
Jody Hagins write: > /hagins@gamecock.rtp.dg.com (Jody Hagins) / 10:55 am Dec 5, 1990 / >Yeh, I wrote one when I was in school too. >I didn't have any problems with it, and as I remember, even >the "slower" students seemed to provide a workable solution. >So, keep working (or should I say start working), and I'm >sure you'll be able to write one soon enough. ... further ranting deleted ... Jody, I feel you're engaging in my favorite exercise: jumping to conclusions bases on little or no evidence. In a free society, people are presumed innocent until proven guilty. Please think before indulging yourself in bad-tempered name calling. I will indulge myself in a little psycho-babble and point out that usually people see themselves in others and judge other's actions on their own wants and desires. In that light, your posting doesn't reflect well on you. Best regards, Jim Bigelow Ft. Collins, CO Advice: The wise don't need it, fools won't head it. All opinions are my own and do not reflect those of Hewlett Packard Inc.
fritz@urz.unibas.ch (12/09/90)
In article <1990Dec7.140547.2976@watserv1.waterloo.edu>, dmurdoch@watserv1.waterloo.edu (D.J. Murdoch - Statistics) writes: > In article <1990Dec6.192941.1215@urz.unibas.ch> fritz@urz.unibas.ch writes: >> >>For a lot of numerical problems I can STRONGLY recommend the book >> >> NUMERICAL RECIPES IN PASCAL -- The Art of Scientific Computing >> >>written by four authors (Press, Flannery et.al.) and published by >>Cambridge University Press (UK). > > I have the original Numerical Recipes, which has the code in Fortran with > fairly literal Pascal translations in an appendix. Have you (or anyone else) > looked at both versions enough to be able to recommend for or against buying > the pure Pascal version? > > Duncan Murdoch To be honest: no ! I also know only the FORTRAN version with the Pascal appendix. I've read in a bookstore that a `Pascal-only' edition has appeared now. I haven't seen it yet, because -- as I was told in the bookstore -- Cambridge is rather slow in deliveries. The C-version of the book is about the same as the FORTRAN-version. So I guess the Pascal-version won't differ to much either. Oliver Fritz
abcscnuk@Twg-S5.uucp (Naoto Kimura (ACM)) (12/09/90)
In article <1990Dec8.182611.1221@urz.unibas.ch> fritz@urz.unibas.ch writes: >In article <1990Dec7.140547.2976@watserv1.waterloo.edu>, dmurdoch@watserv1.waterloo.edu (D.J. Murdoch - Statistics) writes: >> In article <1990Dec6.192941.1215@urz.unibas.ch> fritz@urz.unibas.ch writes: >>> >>>For a lot of numerical problems I can STRONGLY recommend the book >>> >>> NUMERICAL RECIPES IN PASCAL -- The Art of Scientific Computing If this is the book that I think that you're referring to, I wouldn't recommend it for the pascal code, as it is mostly a literal translation of the FORTRAN code (even down to the GOTO's). Although the authors claim that the code was thoroughly tested and assure that it conforms to standard pascal, the code is really only guaranteed to work in Turbo Pascal (there are a lot of assumptions made in the code). The identifiers used throughout the programs are very cryptic, and lots of unused identifiers and labels are strewn throughout the code. I haven't taken a look a the C version of the book, but I suspect it's going to be similar (another literal translation from FORTRAN). Some parts I would disagree with (like claims that the Heapsort is the BEST sort, and sorts like the Shell sort aren't worth discussing). Some sections are pretty good, like the one on random numbers. //-n-\\ Naoto Kimura _____---=======---_____ (abcscnuk@csuna.csun.edu) ====____\ /.. ..\ /____==== // ---\__O__/--- \\ Enterprise... Surrender or we'll \_\ /_/ send back your *&^$% tribbles !!
c145gmk@utarlg.utarl.edu (GORDON KEEGAN) (12/11/90)
In article <4529@idunno.Princeton.EDU>, subbarao@phoenix.Princeton.EDU (Kartik Subbarao) writes... >In article <1990Dec5.175558.29859@dg-rtp.dg.com> hagins@gamecock.rtp.dg.com (Jody Hagins) writes: >>In article <2173@mentor.cc.purdue.edu>, k3u@mentor.cc.purdue.edu (Barry Smith) writes: >>|> Does anyone have the source code to routines that can extract >>|> the complex roots of polynomials? If there is a book that has >>|> the source code that would be fine too. >> >>Yeh, I wrote one when I was in school too. >>I didn't have any problems with it, and as I remember, even >>the "slower" students seemed to provide a workable solution. >>So, keep working (or should I say start working), and I'm >>sure you'll be able to write one soon enough. >> Finding roots for a quadratic was easy enough. However, if you could give me pointers to finding roots to higher degree polynomials, I'd be interested in hearing from you. (no, this is not a lab assignment... :) > >Anyway -- what does everyone else think about this? I've seen both sorts of >attitudes on the subject. What do you guys think comp.lang.c should do? > I won't reply to what is (IMHO) a request for answers to homework. I can't affect what others will do. 'Nuff said. ----------------------------------------------------------------------------- | Gordon Keegan || Bitnet : c145gmk@utarlg | | Systems Programmer || THEnet : UTARLG::C145GMK | | Academic Computing Services || Internet: c145gmk@utarlg.utarl.edu | | University of Texas, Arlington || AT&TNet : 817-273-2208 | ----------------------------------------------------------------------------- | Beauty does what beauty does best: it's beautiful... | -----------------------------------------------------------------------------
ts@uwasa.fi (Timo Salmi) (12/11/90)
In article <10782@helios.TAMU.EDU> c145gmk@utarlg.utarl.edu writes: > > Finding roots for a quadratic was easy enough. However, > if you could give me pointers to finding roots to higher > degree polynomials, I'd be interested in hearing from you. > (no, this is not a lab assignment... :) (Are the postings not getting through since this has already been answered more than once during the last two weeks. Well, no matter, here is an instant replay). If you want the code, take a look at Turbo Pascal Numerical Toolbox, and/or Press & Flannery & al, Numerical Recipes. If you just want a program for solving the roots, get /pc/ts/tsnum12.arc by anonymous ftp from uwasa.fi archives. ................................................................... Prof. Timo Salmi (Moderating at anon. ftp site 128.214.12.3) School of Business Studies, University of Vaasa, SF-65101, Finland Internet: ts@chyde.uwasa.fi Funet: gado::salmi Bitnet: salmi@finfun
capehart@nevada.edu (Anne Racel) (12/12/90)
=> Finding roots for a quadratic was easy enough. However,
=> if you could give me pointers to finding roots to higher
=> degree polynomials, I'd be interested in hearing from you.
=> (no, this is not a lab assignment... :)
What I used, courtesy of my chem instructor (was transferring a program
for pH from Fortran to Pascal, was the Newton-Reismann (sp?) formula.
from calculus. Must admit, there is some bug that hopefully some guru out
there can figure out, that causes the thing to take 20 mintues for a
wrongly rounded anser. But here, essentially, is it:
function NEWTON(TERMS : POLY ;
OLD : real): real;
var
ORDER : integer ; { order of polynomial expression }
H,I : integer ; { counter }
DERIV : real ; { derivative }
Y : real ; { haven't a clue }
CRIT : real ; { criteria for exiting loop }
DUM : real ; { haven't a clue }
NEW : real ; { new value for H+ concentration }
begin
if (TERMS[1]) <> 0 then
begin
Y := 0;
ORDER := 12;
DERIV := 0 ;
repeat { find order of the polynomial }
ORDER := ORDER - 1
until TERMS[ORDER] > 0;
{ end repeat }
if ORDER > 1 then
repeat
for I := 1 to ORDER do {evaluate polynomial}
begin
Y:= Y + TERMS[I] * EXP(I * LN(OLD));
DERIV := DERIV + ((I-1) * TERMS[I] * EXP((I-2)*LN(OLD)));
end;
NEW := OLD - (Y/DERIV);
CRIT := ABS(1.0e6 * ((OLD-NEW)/OLD)); {check whether old value and new value close }
OLD := NEW;
Y := 0;
DERIV := 0
until CRIT <= 1
else
NEW := OLD
end;
NEWTON := NEW
end;
..... and for those familiar with Fortran, here's that version, which DOES
work:
$STORAGE:2
SUBROUTINE NEWTON (POLY,X,NEWX)
INTEGER H, I
INTEGER*4 CRIT
REAL*8 POLY, DERIV, X, Y, NEWX, DUM
DIMENSION POLY(11)
C FIND THE ORDER OF THE POLYNOMIAL EXPRESSION STORED IN THE MATRIX
DO 5, H=11, 1, -1
IF (POLY(H)) 10, 5, 10
5 CONTINUE
10 DO 20, I=1, H
Y = Y + POLY(I)*X**(I-1)
IF (I.LE.1) GOTO 20
DERIV = DERIV + (I-1)*POLY(I)*X**(I-2)
20 CONTINUE
NEWX = X - (Y/DERIV)
DUM = 1000000.*((X-NEWX)/X)
CRIT = ABS(INT(DUM))
IF (CRIT.LE.1) GOTO 30
X = NEWX
Y = 0
DERIV = 0
GOTO 10
30 RETURN
END
=============================================================
Anne Racel Internet: capehart@nevada.edu
University of Nevada, Compuserve: 72105,1105
Las Vegas Bitnet: capehart@unsvax.bitnet
"I'm very certain, Oz, that you gave me the best brains in the world
for I can think with them day and night, when all over brains are
fast asleep."
Scarecrow in "Dorothy and the Wizard in Oz"