Discussion Overview
The discussion revolves around methods for estimating a function that fits well with a given curve, focusing on curve fitting techniques and software options. Participants explore various mathematical approaches and tools for achieving better curve fitting, particularly in the context of data with significant variation.
Discussion Character
- Exploratory, Technical explanation, Debate/contested
Main Points Raised
- One participant seeks advice on how to derive a function that fits a specific curve, sharing a link to the curve image.
- Another participant suggests the LOESS method as a potential solution for curve fitting.
- A participant mentions successfully using a fourth-degree polynomial in Excel for fitting but expresses concern about its effectiveness with more variable graphs, indicating a search for software that employs alternative methods like moving averages or exponential functions.
- One participant critiques the use of polynomials for fitting, particularly in regions of steep slopes, and suggests that coordinate transformations or LOESS might yield better results.
- A participant shares a link to an Excel add-in that claims to have LOESS capabilities and proposes considering parametric cubic polynomials, such as Bezier curves, for handling vertical slopes.
- A later reply expresses gratitude for the shared information, noting that it provided a very good fitting solution.
Areas of Agreement / Disagreement
Participants express differing opinions on the effectiveness of polynomial functions for curve fitting, with some advocating for alternative methods like LOESS or parametric cubic polynomials. The discussion does not reach a consensus on the best approach.
Contextual Notes
Participants highlight limitations of polynomial fitting in certain regions of the curve and the potential need for coordinate transformations or alternative fitting methods, but do not resolve these issues.
