Abstract
This paper presents a software implementation of an algorithm for processing experimental data obtained using the probabilistic-deterministic design of experiments (PDDoE) method in the R environment. The tool enables the construction of partial and generalized dependencies between results and varying factors, supports templates of different dimensionalities, automates the selection of approximation models, and evaluates their quality using R², the nonlinear multiple correlation coefficient (RM), and its significance (tRM). Implementation is carried out entirely in R using packages such as openxlsx, dplyr, and tcltk. The system supports arithmetic, geometric, and harmonic averaging methods, automatic orthogonality checks, and prioritized model selection based on tRM, with emphasis on physically meaningful functions. Generalized equations are constructed by enumerating combinations of significant partial dependencies until the maximum tRM is achieved. The tool provides result visualization, table export, and a modular structure allowing easy extension. Users can load custom plan templates, add new metrics and approximating functions, and manually define the form of both partial and generalized models. Designed for researchers, educators, and engineers working with multifactorial systems, the tool is applicable in scientific research, chemical technology, and education. The ChatGPT language model was used in R code generation. Potential applications of AI tools in experimental data processing are briefly discussed.