PPlot, a webapp to partition geochemical data and isolate mixed subpopulations using probability plot modeling





geochemical mapping, multimodal populations, mixed subpopulations, isolating subpopulations, probability plot


Statistical methods are mostly designed to handle datasets comprising statistically single normal or log-normal populations, but geochemical and geophysical surveys usually deviate from this expectation. A reason for this is the heterogeneity in the occurrence of geological objects, so the complete dataset may correspond to multiple mixed subpopulations. Specifically, multiple mixed subpopulations can refer to differences between mineralized and barren areas, different geochemical facies of a geological unit, or contaminated and healthy areas. This implies a restriction on using classical or even robust statistical estimates, unless the underlying subpopulations can be extracted from the dataset. The probability plot can be used to assess a dataset and to infer a possible combination of subpopulations, either normal or log-normal, whose combination may generate it. The web-based app PPlot, presented in this paper, allows the plotting of the probability plot of a dataset and modeling the underlying subpopulations present in it, either automatically or manually. After modeling the dataset by the application, the user will obtain numerical results and plots of the range of values that delimit each subpopulation, as well as the mean and standard deviation for each of them. Computer-generated and real datasets were used to validate the procedure and coding, and an example of usage is presented. The app was developed using HTML5 and JavaScript and it runs in any modern browser, and is freely available in https://pplotweb.firebaseapp.com/.


Download data is not yet available.


Additional Files



How to Cite

Ferreira de Campos, F., Licht, O. A. B., & Campos, N. B. F. (2023). PPlot, a webapp to partition geochemical data and isolate mixed subpopulations using probability plot modeling. Geochimica Brasiliensis, 37, e–23002. https://doi.org/10.21715/GB2358-2812.202337002