FittingKVdm

Working with data of various origins, in a scientific context that involves plotting and pattern identification can pose multiple challenges, ranging from the ones that are related to its specificity, all the way to those who address more general aspects, such as distribution, outliers, bias, and more.

FittingKVdm, is an application that was constructed specifically in order to offer users a specialized tool for identifying patterns and trends in their data, regardless of its nature and characteristics.

The application relies on an iterative process, which can be applied to a multitude of functions, therefore eliminating the requirements of variable logs functions for increasing precision. The application accounts for the assigned data point weights, with standard deviations for both x and y, which are also employed in order to perform a parametric estimation of the resulting confidence intervals.

There are a number of models available for selection, which were included for educational purposes, such as constant, linear, quadratic, and much more, which have been tested by the developer with real-world data.

Last but not least, one unique feature will enable users, in the case of invertible model functions, to perform the fitting in any x or y directions, therefore yielding direct improvements.