BatchGraphPlotter.py

import pandas as pd
import matplotlib.pyplot as plt
from matplotlib.pyplot import cm
import csv
import numpy as np
import math

numberOfDifferentK = 10
parent='out/e/'
with open('Elbow.csv', 'w', encoding='UTF8', newline='') as f:
    writer = csv.writer(f)
    for counter in range(1, numberOfDifferentK+1):
        data = []
        cen = pd.read_csv(parent+str(counter)+'/Centroids.csv', header=None)
        k = len(cen)
        centroids = []
        p = []
        ss = []
        for i in range(k):
            p.append([[], []])
            ss.append(0)
        with open(parent+str(counter)+'/Centroids.csv', 'r') as cf:
            reader = csv.reader(cf, delimiter=',')
            for row in reader:
                centroids.append([float(row[0]), float(row[1])])
        n = [0]*k
        with open(parent+str(counter)+'/Output.csv', 'r') as f:
            reader = csv.reader(f, delimiter=',')
            for row in reader:
                for i in range(k):
                    x = float(row[0])
                    y = float(row[1])
                    x1 = float(row[2])
                    y1 = float(row[3])
                    if(x == centroids[i][0] and y == centroids[i][1]):
                        n[i] = n[i]+1
                        p[i][0].append(x1)
                        p[i][1].append(y1)
                        d1 = (x-x1)
                        d2 = (y-y1)
                        ss[i] = ss[i]+d1*d1+d2*d2
                        break
        wcss = sum(ss)
        data.append(k)
        data.append(wcss)
        for i in range(k):
            ss[i] =  math.sqrt(ss[i]/n[i])
        #Mean Distortion
        data.append(sum(ss)/k)
        writer.writerow(data)
        # plotting points as a scatter plot
        color = iter(cm.rainbow(np.linspace(0, 1, k)))
        for i in range(k):
            c = next(color)
            ax = plt.gca()
            ax.add_patch(plt.Circle(
                (centroids[i][0],  centroids[i][1]), ss[i], facecolor='#FF000000', edgecolor=c))
            ax.set_aspect('equal')
            plt.scatter(p[i][0], p[i][1], color=c, marker=".", s=30)
            plt.scatter(centroids[i][0], centroids[i][1],
                        label="Centroid " + str(i+1), color=c, marker="*", s=60)

        # x-axis label
        plt.xlabel('X (cm) - >')
        # frequency label
        plt.ylabel('Y (kg) - >')
        # plot title
        plt.title('K-Means Clustering Output')
        # showing legend
        plt.legend()

        # function to show the plot
        # plt.show()