New Trapedoizal Tail integration method
05-14-2021, 03:21 AM
Post: #7
 Namir Senior Member Posts: 1,083 Joined: Dec 2013
RE: New Trapedoizal Tail integration method
(05-13-2021 12:05 PM)Albert Chan Wrote:
(05-12-2021 06:05 PM)Namir Wrote:  The method samples two function values in each interval at f(x+h/2-h/3) and fx(x+h/2+h/3)

You might also try f(x+h/2 ± h/(2√3))
This is the points and weights for 2 points Gaussian quadrature.

Because 2√3 ≈ 3.464 relatively close to 3, this explained the good result you observed.

Thanks Albert! The 2*sqrt(3) makes sense. I tried several points on either side of the half interval with varying degrees of success.

Namir
 « Next Oldest | Next Newest »

 Messages In This Thread New Trapedoizal Tail integration method - Namir - 05-12-2021, 06:05 PM RE: New Trapedoizal Tail integration method - Paul Dale - 05-13-2021, 06:17 AM RE: New Trapedoizal Tail integration method - Namir - 05-13-2021, 08:28 AM RE: New Trapedoizal Tail integration method - Maximilian Hohmann - 05-13-2021, 12:12 PM RE: New Trapedoizal Tail integration method - Albert Chan - 05-13-2021, 01:41 PM RE: New Trapedoizal Tail integration method - Albert Chan - 05-13-2021, 12:05 PM RE: New Trapedoizal Tail integration method - Namir - 05-14-2021 03:21 AM

User(s) browsing this thread: 1 Guest(s)