From: Jean-Yves Avenard on 4 Sep 2006 00:18 Michael Kuyumcu wrote:> Arbitrary precision numbers on the HP? I am aware only of *integers* > with multiple precision. Why didn't HP follow through on implementing > multiple-precision arithmetic also for floats? Now that the basic There's full support in the OS for almost unlimited precision numbers. up to 524,288 digits with an exponent of up to 524,288 digits Of course you can't ever create such number due to the RAM limitation. JY From: Paul Schlyter on 4 Sep 2006 03:43 In article , Steen Schmidt wrote: > Paul Schlyter wrote: > >>> You can do games and such, but when you need to do some data IO with >>> the OS (even a plotter app does that), you're bound by the >>> limitations of the existing OS (data types, argument ranges, file >>> structure etc). >> >> If you do C/C++ without any OS, you have no such limitations. I mean, >> in such a situation there is no OS; so how could a non-existent OS >> limit you? > > What would you do when you have calculated 10000! ? Would you display > the 35660 digits on the screen for the user to decipher one at a time? :-) ...an interesting suggestion. If one new digit is displayed each second, it would take almost 10 hours to display the result, since: 10000! = 2.846259680917054518906413212119868890148051401702799230794179 99427441134000376444377299078675778477581588406214231752883004233994015 3518739052421161383E+35659 ...approximately > What if the result was to be used by a subsequent calculation? Or let's > say you coded a new symbolic integration engine in C - how would you > output the integral of 'Exp(X^2),X'? By turning on the appropriate pixels in the display, of course! > The result is > '1/2*Sqrt(pi)*Erfi(X)', but Erfi(X) isn't defined in the TI AMS. You > could code Erfi(X) in C, but the result had to reside temporarily in > the TI OS environment, WHAT "TI OS" ???? I was discussing a C or C++ program without any OS.... > until the user decides to do something else with > the result (evaluate it numerically or integrate it again etc.). What > object type would Erfi(X) be in this case, inside the TI AMS? > > The OS (and its aux sw like parsers, type checkers, CAS etc.) limits > you in what you can return to it or get as arguments from it. On the HP > we have built-in data types for arbitrary precision numbers, arrays of > any type etc. There are libraries with support for user-defined > functions that behave exactly like built-in functions. This is not the > case on the TI - not even on the NSpire. The latter I find > disappointing, as it would be a good time to step up from the silly > limitations of the TI92/92+/V200/89/89ti series. > > But TI will probably not do this, as the open structure of the HP OS > opens up for a hornets nest of bug possibilities. If you look at how > flexible the HP OS is - how much is available for the user - then it's > obvious that there's much greater risk of bugs existing therein in > comparison with the closed TI OS. > > All the above is moot of course, in the event you'd want to code your > own OS in C/C++. And even in that case, it would probably have to be > done over a couple of times, as you'd probably not succeed in putting > in all the features another programmer wished for, in which case that > other programmer had to code his or her own special OS etc... Coding without any OS support is indeed a lot of work - if your program is non-trivial. > Regards > Steen -- ---------------------------------------------------------------- Paul Schlyter, Grev Turegatan 40, SE-114 38 Stockholm, SWEDEN e-mail: pausch at stockholm dot bostream dot se WWW: http://stjarnhimlen.se/ From: Michael Kuyumcu on 4 Sep 2006 07:44 The same result in approx mode, but what a difference in speed! Is this buggy, or what? Regards, Michael Kuyumcu Yao Konan schrieb: > Hi Michael, > > Have you tried in approx mode ? > Because on my TI92+ both matrix(especially the first one) don't look > singular in approx mode. > The determinant computations take too much time. > > Regards, > Konan Yao > > Michael Kuyumcu a écrit : > > > Hi Yao, > > > > I just tested: > > o "seq(seq(1/(sqrt(2*i+1-I*(3*j+2))),i,1,3),j,1,3)^-1", which yields > > "Singular matrix", and then > > o "seq(seq(1/(sqrt(2*i+1)+I*sqrt(3*j+2)),i,1,3),j,1,3)^-1", and again: > > "Singular matrix". > > I havent's double-checked with the hp49g+ whether or not this matrix is > > in fact singular, but if not, we might have hit on a CAS error. > > > > Regards, > > Michael > > > > From: Michael Kuyumcu on 4 Sep 2006 07:46 Hey, that's terrific! Thanks for the info. Will I get the longfloat info at hpcalc.org? Great! Regards, Michael Kuyumcu Steen Schmidt wrote:> Michael Kuyumcu wrote: > > > Arbitrary precision numbers on the HP? I am aware only of integers > > with multiple precision. Why didn't HP follow through on implementing > > multiple-precision arithmetic also for floats? > > The data type is built-in for arbitrary floating point real and complex > numbers (type 27). If you install the Longfloat library you also have > arbitrary floating point arithmetic at your fingertips (using that > built-in datatype). Pi to 200 decimal places in 2.5 seconds for > instance. And that is not in C. > > Regards > Steen From: Steen Schmidt on 4 Sep 2006 12:58 Michael Kuyumcu wrote: > Hey, that's terrific! Thanks for the info. Will I get the longfloat > info at hpcalc.org? Yes, you can find the latest longfloat lib here: http://www.hpcalc.org/details.php?id=5363 Regards Steen First  |  Prev  |  Next  |  Last Pages: 10 11 12 13 14 15 16 17 18 19 20 21 22 Prev: HP50g vs. Voyage 200Next: HPUserEdit crash at launch