NavList:
A Community Devoted to the Preservation and Practice of Celestial Navigation and Other Methods of Traditional Wayfinding
Re: Linear regression and other tools
From: Brad Morris
Date: 2018 Oct 15, 12:58 -0400
From: Brad Morris
Date: 2018 Oct 15, 12:58 -0400
Tony
Many calculators include a "random number" function.
Decide on the greatest error you wish to include, and then, with a bit of arithmetic, you can randomize your noiseless data.
Example: calculator random number is from 0 to 1.0.
Subtract 0.5 from random number to yield -0.5 to 0.5
Multiply by 2, yielding -1.0 to 1.0
Multiply by Greatest Error.
Add to noiseless data.
In spreadsheets, you can sometimes also include the distribution of those random numbers. The process above yields a uniform distribution (any error is equally as likely as any other).
Brad
On Mon, Oct 15, 2018, 12:34 PM Tony Oz <NoReply_TonyOz@fer3.com> wrote:
Dear Antoine,
I will certainly re-do the simulations as you asked.
Regarding the total lack of any noise in my previous example: yes, I clearly do understand this. I just could not quickly invent a way to introduce such noise. My next attempt will be a truncation of HCs to the integer arcminutes (but I not sure if this would be enough).
So far I'm really impressed with TI-83+'s capabilities, and if a calculator is available - one should use it to its' full powers.
Warm regards,
Tony
60°N 30°E
PS
It's nice Indian Summer here (60°N 30°E ), warm and quiet. :)
