Global Journal of Human-Social Science, B: Geography, Environmental Science and Disaster Management, Volume 22 Issue 3

Table 3: Models used to express the resilient behavior and permanent deformation of the soils of this research References Equation Variables Parameters Hicks and Monismith (1970) MR=k13k2 3 k1 and k2 Svenson (1980) MR=k1dk2 d k1 and k2 Macêdo (1996) MR=k13k2dk3 3;d k1, k2 and k3 Guimarães (2009) p(%)=1302d03N4 3;d and N 1, Ψ 2, 3 and 4 Notation: 3 = confining stress; d= deviator stress; p(%) = Specific permanent deformation, N = number of loading cycles and, k1, k2, k3, 1, Ψ 2, 3 and 4 are regression parameters, 0 = reference stress, considered with the atmospheric pressure of 100 kPa. c) Data Science Tools For Cluster Analysis The implementation of the codes required for the analyses was carried out in the Python programming language (version 3) available in the Anaconda virtual environment (https://www.anaconda.com/) . In addition to the standard Python library, modules and functions from other libraries were used as indicated in Table 4, which explains the general objectives of each (in the application column) and in the column "functions and modules" is the specific indication of the main tools that were used. Table 4: Main Libraries, modules and functions used in the analysis with Python in this research Libraies Aplication Modules and functions Pandas https://pandas.pydata.org/ Data manipulation and analysis: structures and operations to manipulate numerical tables and time series. read_excel, drop, set_index, head, shape Numpy https://numpy.org/ Data analysis: mathematical functions, random number generators, linear algebra routines, Fourier transformations, etc. Array, arrange, mean, std, argsort Scipy https://www.scipy.org/ Data modeling: fundamental algorithms for statistical functions (probability distributions, hypothesis testing, frequency statistics, correlation functions, etc.). scipy.cluster, hierarchy, hierarchy.linkage (method='ward',metric='euclidean'), hierarchy.dendrogram Scikit-learn https://scikit-learn.org/ Data Modeling: Machine learning algorithms, supervised and unsupervised (Classification, Regression, Clustering, Model Selection, etc.) sklearn.preprocessing, StandardScaler, transform Scikit-image https://scikit-image.org/ Image processing: functions for manipulating scientific, specific or general-purpose images, operations on Numpy matrices, manipulation of exposure and color channels, detection and segmentation of objects. Feature Imageio https://pypi.org/project/imageio/ Image manipulation: reading and writing image data, including animated images, volumetric data, and scientific formats. imread Matplotlib https://matplotlib.org/ Data presentation and exploration: creating graphs and general data visualization plot.figure, plot.title, plot.xlabel, , plot.yabel, subplot , imshow For the analysis of data clusters through the hierarchical method and implementation via codes and libraries in Python, the data were standardized by applying the Z-score Normalization Method, and for this an array was created through the "array" function of the Numpy library and the functions "StandardScaler" and "transform" of the Scikit-learn library were applied. Then, to obtain hierarchical cluster of the (standardized) datasets, the hierarchy function of the Cluster module of the SciPy library was imported. The implementation was made considering Ward's linkage method and Euclidean distance. Volume XXII Issue III Version I 14 ( ) Global Journal of Human Social Science - Year 2022 © 2022 Global Journals B Clustering of Fine-Grained Tropical Soils using Data Science Tools Applied to their Geotechnical Properties