Hi all!
I'm trying to solve a problem about a circle inscribed in a triangle using Newton's method. The Root function does not work.
I have some questions with how muted operators work - I don't want to set the value n - I want to use a hybrid operators for and while.
What am I doing wrong?
Triangle-Circle-Newton.sm (143 КиБ) скачан 31 раз(а).

See please more https://community.ptc.com/t5/Mathcad/Malfatti-circles/td-p/687281

https://community.ptc.com/t5/image/serverpage/image-id/31378i6EB179E7EF72B74C/image-size/large?v=v2&px=999

Hi, first of all you have to keep hands off i if you handle complex numbers. Don't use it as iterator variable.

Second, I am not sure what for(4) actually does. Unfortunately I can't break down it to a minimum working example. The disabled region should work as the for(3) loop with exit condition but it doesn't (somehow gets stuck). Tests with trace showed that inside the for(4) loop the arithmetics seem to be rotten. For comparison I added a loop-free implementation of the algorithm by just repeating the commands in the sheet.

Once one accepts the complex numbers, the algorithm converges to a reasonable solution. Yet, the radius turns out negative for the given initial guess. This can happen because only the square of the radius occures in the model.


Triangle-Circle-Newton_kr.sm (194 КиБ) скачан 27 раз(а).

Wrote

What am I doing wrong?

Augmented version of the original MCD 11
Jean

Maths Triangle ScaleneInscribed.sm (41 КиБ) скачан 29 раз(а).

I don't think you've actually adopted the KISS principle here Valery! The attached seems much simpler.


Triangle-Circle-Newton_stvmath.sm (116 КиБ) скачан 45 раз(а).