The article describes the stages and results of the multidimensional classification of personal subsidiary plots (PSP) according to the All-Russian Agricultural Census of 2016 (VSHP-2016). This classification was carried out on the basis of four prepared sets of initial data, two of which are synthetic indicators in the form of multidimensional means, and implemented using the Python programming language, including the skit-learn library for clustering and the matplotlib library for graphical visualization of the resulting split of households into homogeneous groups. The original data is normalized using the L2 Normalization method, also known as Spatial Sign Preprocessing. Cluster analysis is carried out using the k-means method based on the Lloyd algorithm, the number of clusters is determined using the Silhouette Coefficient. The results are visualized through bar charts, 2D and 3D scatter plots.