HP Forum Archive 17

[ Return to Index | Top of Index ]

a 32sii equation bug? Message #1 Posted by Les Wright on 17 Mar 2007, 8:23 a.m. A recent thread dealt with integrating the bell curve integrand, exp(-x^2), on the HP48G, and I have been playing with the problem on my recently acquired 32sii. When trying to integrate this expression as an equation, the calculator does not seem to like exp(-x^2) in that very form and never seems to converge. But it seems to like exp(-(x^2)), 1/exp(x^2), and exp(-sq(x)) just fine. It also works adequately when the function is provided as an RPN program (RCL X, x^2, +/-, e^x). The 33S handles exp(-x^2) just fine in that form, without any need for extra parentheses or alternate representations of the quantity to be antilogged. Weird.... Les

Re: a 32sii equation bug? Message #2 Posted by Thomas Radtke on 17 Mar 2007, 8:52 a.m., in response to message #1 by Les Wright In this case, the 32sii can't tell minus from a signed numer, which has a higher operator precedence as the exponentiation. So, +/- seems to be the same token as '-' in the equation editor. If you write explicitly exp(0-x^2) instead, everything works fine. Edit: And congrats to the 32sii! :-) Edited: 17 Mar 2007, 8:54 a.m.

Re: a 32sii equation bug? Message #3 Posted by Les Wright on 18 Mar 2007, 3:32 p.m., in response to message #2 by Thomas Radtke Quote: And congrats to the 32sii! :-) I am particularly proud of this acquistion. It is in impeccable shape, and I got it for $130 from a fellow Canadian, so no customs and shipping was inexpensive and quick. I didn't think this was such a big deal, but then I learned that some folks pay twice that for this item. As a lover of the 33S, I am right at home with it, though the tiny memory size takes some adjustment. It came with a nice case too--the thicker lined stitched leatherette one, not the thin molded vinyl one. Les

Re: a 32sii equation bug? Message #4 Posted by bill platt on 17 Mar 2007, 6:54 p.m., in response to message #1 by Les Wright It isn't a bug as much as a poorly implemented system for "unary minus" that is at the root of the problem. There is a "bug" aspect to the unary minus with the 32sii but that is not really the point. If you read the manual for the 32sii, and then read the 33s manual, you will see that in fact the "unary minus" function was changed and improved (simplified) in the 33s. The 32sii has a bad behavior with it in equations and reading the manual, you can see that it might be trouble. I think Jordi Hidalgo's DATAFILE article on the 33s has some discussion of this, too. As Thomas points out, just use the parenthesis whenever you have leading minuses, and it makes life easier in the 32sii.

Re: a 32sii equation bug? Message #5 Posted by Les Wright on 18 Mar 2007, 3:26 p.m., in response to message #4 by bill platt Quote: If you read the manual for the 32sii, and then read the 33s manual, you will see that in fact the "unary minus" function was changed and improved (simplified) in the 33s. The 32sii has a bad behavior with it in equations and reading the manual, you can see that it might be trouble. So the much-denigrated 33S actually IMPROVED on something in a pre-Kinpo calculator? And a Pioneer at that? Egad, the scandal! Les

Re: a 32sii equation bug? Message #6 Posted by Rav on 19 Mar 2007, 12:51 p.m., in response to message #5 by Les Wright Actually I think that the 33s improved on the equation solver too. It seems that the 33sii is an iterative solver and so can fall over. The people who kindly responded to this question, http://www.hpmuseum.org/cgi-sys/cgiwrap/hpmuseum/archv015.cgi?read=85299 explain it much more eloquently. Rav

[ Return to Index | Top of Index ]

Go back to the main exhibit hall