Best fit funtion reprenting raw data points

02072014, 06:19 AM
Post: #1




Best fit funtion reprenting raw data points
If you have a set of raw data points (x;y), how do you enter the data points on the HP Prime and fit a graph/funtion to the data points, eg a best fit quadratic or linear function to represent the data points? Cant find it in the manual.


02072014, 01:55 PM
Post: #2




RE: Best fit funtion reprenting raw data points
Use the statistics 2 var app, enter the data in NUM, SYMB to setup your regressions, PLOT to display/calculate.
The QSG that comes with the calculator has 2 pages describing this (p3738) and the full manual has complete info (p223238). Note that is in the english pagination. TW Although I work for the HP calculator group, the views and opinions I post here are my own. 

02072014, 07:18 PM
Post: #3




RE: Best fit funtion reprenting raw data points
(02072014 01:55 PM)Tim Wessman Wrote: Use the statistics 2 var app, enter the data in NUM, SYMB to setup your regressions, PLOT to display/calculate. Thank you Tim. You are always so helpful. I hope the HP bosses notice and give you a big bonus :) 

08292017, 08:01 AM
Post: #4




RE: Best fit funtion reprenting raw data points
Refloating this question for a deeper answer.
Is there any command that analyzes all types of curves for getting the best fit (best correlative factor) without manual searching? I know HP48 had this function. Thank you. 

08292017, 08:43 AM
Post: #5




RE: Best fit funtion reprenting raw data points
There was some time ago a detailed thread discussion on this topic, including a BestFit function for linear models (post #12).


08292017, 09:43 AM
Post: #6




RE: Best fit funtion reprenting raw data points
Interesting link. Thank you for it.


08292017, 12:27 PM
Post: #7




RE: Best fit funtion reprenting raw data points
I just read that thread and was surprised that nobody suggested doing a Lagrange interpolation of all the points to get "the best fit" :)
In my opinion Lagrange interpolation clearly shows how meaningless the idea of "the best fit" is if you do not restrict the domain of your models (and even if you do not restrict it enough, e.g. to all polynomials). Especially when you consider that you can add any number of randomly generated additional points to an existing data set (as long as all of the x values are different) and still always produce a perfect fit. 

« Next Oldest  Next Newest »

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