HP Forum Archive 18

[ Return to Index | Top of Index ]

Calculator forensics Message #1 Posted by designnut on 5 June 2008, 8:27 p.m. I read of the accuracy test performed using 9 sin cos tan arc tan arc cos arc sin in sequence. In the 50g in the degrees mode it shows 8.999999864267. I wondered how it would perform in the radians mode so I used Pi/20 (same 9 degrees) repeated the test and divided the results into the original angle and multiplied by 9 to get a comparable number (9). It showed 8.999999999995 It seems the algorithms are more compatible with the radian mode. Sam

Re: Calculator forensics Message #2 Posted by Karl Schneider on 6 June 2008, 1:25 a.m., in response to message #1 by designnut Quote: It seems the algorithms are more compatible with the radian mode Hi, Sam -- Have a look here (discussion from me): http://www.hpmuseum.org/cgi-sys/cgiwrap/hpmuseum/archv017.cgi?read=113143#113143 And here (discussion from Valentin): http://www.hpmuseum.org/cgi-sys/cgiwrap/hpmuseum/archv014.cgi?read=69026 In fairness, the author of the forensic did not intend it to be a measure of a calculator's general accuracy; he intended it to be an indicator of internal algorithms and precision for purposes of comparison between models. -- KS

[ Return to Index | Top of Index ]

Go back to the main exhibit hall