HP Forums / HP Museum Forums / Old HP Forum Archives v / WP34s quadrature

Threaded Mode
WP34s quadrature
Post: #1
04-26-2011, 05:52 AM
hugh steers Offline
Unregistered
Posts: 536
Threads: 56
Joined: Jul 2005

Hi,

the manual says this works like the 15c, but i get no difference when changing (eg) FIX mode size. is there some way of specifying the accuracy - or am i doing something wrong.

right now, im getting really appalling results. 2 figures correct, for example.

Find
Post: #2
04-26-2011, 05:54 AM
Paul Dale Offline
Unregistered
Posts: 3,229
Threads: 42
Joined: Jul 2006

There is no way to change the precision of the integration.

We're using a Gauss/Kronrod quadrature with a fixed number of points. It won't be terribly good for "difficult" functions.


- Pauli

Find
Post: #3
04-26-2011, 06:40 AM
Walter B Offline
Unregistered
Posts: 4,587
Threads: 105
Joined: Jul 2005

The manual clearly says "the user interface is as in HP-15C." Nothing more.

Walter

Find
Post: #4
04-26-2011, 06:56 AM
Marcus von Cube, Germany Offline
Unregistered
Posts: 3,283
Threads: 104
Joined: Jul 2005

We can add a more sophisticated approach as a user code routine. Any contributions are welcome.

Find
Post: #5
04-26-2011, 07:59 AM
Namir Offline
Unregistered
Posts: 2,247
Threads: 200
Joined: Jun 2005

If I am not mistaken, the SOLVE and INTEGRATE functions had Professor Kahan as it's mentor. Kahan's philosophy was to push the accuracy to the machine limits and not by a user-defined value.

The HP-15C Advanced Functions Handbook has a section "Accuracy of the Function to be Integrated" that specifies the accuracy of the integral, which is basically the accuracy of the function you are integrating. This tells me that if you are integrating a function using an empirical approximation, then expect more error in the integral.

Namir

Edited: 26 Apr 2011, 8:35 a.m.

Find


Forum Jump: