﻿readme of results:
===============================================================
In the "ClusterPictures" directory, there are 2 subdirectories:
"randomDatasets" - a directory containing the results of 
clustering datasets generated using a randomizer.
"byPictures" - a directory containing the results of clustering 
datasets generated based on pictures.

===============================================================
Files with the extension *.xlsx are tables with numerical 
reports. The workbook itself consists of report pages for 
clustering methods and the final resulting table, which 
includes average values for each method and their comparisons.
* When dividing the network into a number of clusters equal to 
a power of two, clustering methods are more. *
---------------------------------------------------------------
Each page has a number of parameters:
"Nodes Count" - the number of nodes in the entire simulated 
network.
"Clusters count" - the number of clusters/subnets to which 
the target network is divided.
"Maximum validators on Node" - the maximum number of validators 
on a single node.
"Maximum delay" - the maximum delay value in the network.
"Name of algorithm" - the name of the algorithm.
"Algorithm's params" - the parameters of the algorithm.

Next, there is a table with the following columns:
"Dataset Name" - the name of the dataset.
"Cluster Х" - the cluster number.
"Weight_diwiation" - deviation from the average ideal weight.
"Sil" - the silhouette value of the network.
"I_diam" - the diameter value of the network.
"I_sq" - the squareness value of the network.
"Spherical" - the sphericity metric of the network.
The last row represents the average value across all rows. 
On the last sheet, the resulting table has similar parameters, 
except that each row has a cell with the name of the algorithm 
and a cell with the algorithm's parameters instead of the 
dataset name.
---------------------------------------------------------------
Additionally, this folder contains folders in the "CL-N" format, 
which store figures with clustering results, where N is the 
number of clusters to which the network is divided. Each folder 
contains images in the "*.bmp" format that visualize the work 
of the algorithms. The name of each algorithm consists of the 
name of the clustering algorithm and its passed parameters.
===============================================================
In addition to everything described above, the "byPictures" 
directory contains "*.png" images that serve as patterns for 
constructing datasets.
===============================================================