Supplementary data: Applicability of Random Simulation to the Evaluation of Interpolation (Sode, 2024).
GitHub: https://github.com/YujiSODE/Sode2024_AnnualMeeting_PalaeontSocJp-supplementaryData
Copyright (c) 2024 Yuji SODE <yuji.sode@gmail.com>
Results are based on results by MCSS.
R* = ln(1+dL/dx), where dx = max(x*)-min(x*) and dL = (maximum length)-(minimum length) = L1+L2+L3-max(L)
Figure 1. Graphs showing frequency distributions of ln(1+dL/dx).
This graph shows frequency distributions of ln(1+dL/dx), which are defined using three random points on a given function f(x) or generated sample data.
Every line is result of given function f(x). Every point is result of generated sample data. Each graph represents average of a hundred sets of results.
Filled areas or error bars represent standard deviations.
sample data: 100 points
f(x)*Q(0.2): f(x) = (3x-2)^2, Q(p) = 1+p(2u-1), u = [0, 1], x = 2 to 5
Figure. 2. Graphs showing sample data and results of curve fitting.
Every point is plot of the generated data set from a model which is (3x-2)^2 with 20 percent of deviation.
Every line is result of curve fitting using the least squares method.
- Sode, Y., 2024: Applicability of Random Simulation to the Evaluation of Interpolation. on the 2024 Annual Meeting. Palaeontological Society of Japan, Abstracts with Programs, P72-C, p. 53.