Added: clustering algorithm

Klustering algo/Mall_Customers.csv

@@ -0,0 +1,201 @@
+CustomerID,Genre,Age,Annual Income (k$),Spending Score (1-100)
+0001,Male,19,15,39
+0002,Male,21,15,81
+0003,Female,20,16,6
+0004,Female,23,16,77
+0005,Female,31,17,40
+0006,Female,22,17,76
+0007,Female,35,18,6
+0008,Female,23,18,94
+0009,Male,64,19,3
+0010,Female,30,19,72
+0011,Male,67,19,14
+0012,Female,35,19,99
+0013,Female,58,20,15
+0014,Female,24,20,77
+0015,Male,37,20,13
+0016,Male,22,20,79
+0017,Female,35,21,35
+0018,Male,20,21,66
+0019,Male,52,23,29
+0020,Female,35,23,98
+0021,Male,35,24,35
+0022,Male,25,24,73
+0023,Female,46,25,5
+0024,Male,31,25,73
+0025,Female,54,28,14
+0026,Male,29,28,82
+0027,Female,45,28,32
+0028,Male,35,28,61
+0029,Female,40,29,31
+0030,Female,23,29,87
+0031,Male,60,30,4
+0032,Female,21,30,73
+0033,Male,53,33,4
+0034,Male,18,33,92
+0035,Female,49,33,14
+0036,Female,21,33,81
+0037,Female,42,34,17
+0038,Female,30,34,73
+0039,Female,36,37,26
+0040,Female,20,37,75
+0041,Female,65,38,35
+0042,Male,24,38,92
+0043,Male,48,39,36
+0044,Female,31,39,61
+0045,Female,49,39,28
+0046,Female,24,39,65
+0047,Female,50,40,55
+0048,Female,27,40,47
+0049,Female,29,40,42
+0050,Female,31,40,42
+0051,Female,49,42,52
+0052,Male,33,42,60
+0053,Female,31,43,54
+0054,Male,59,43,60
+0055,Female,50,43,45
+0056,Male,47,43,41
+0057,Female,51,44,50
+0058,Male,69,44,46
+0059,Female,27,46,51
+0060,Male,53,46,46
+0061,Male,70,46,56
+0062,Male,19,46,55
+0063,Female,67,47,52
+0064,Female,54,47,59
+0065,Male,63,48,51
+0066,Male,18,48,59
+0067,Female,43,48,50
+0068,Female,68,48,48
+0069,Male,19,48,59
+0070,Female,32,48,47
+0071,Male,70,49,55
+0072,Female,47,49,42
+0073,Female,60,50,49
+0074,Female,60,50,56
+0075,Male,59,54,47
+0076,Male,26,54,54
+0077,Female,45,54,53
+0078,Male,40,54,48
+0079,Female,23,54,52
+0080,Female,49,54,42
+0081,Male,57,54,51
+0082,Male,38,54,55
+0083,Male,67,54,41
+0084,Female,46,54,44
+0085,Female,21,54,57
+0086,Male,48,54,46
+0087,Female,55,57,58
+0088,Female,22,57,55
+0089,Female,34,58,60
+0090,Female,50,58,46
+0091,Female,68,59,55
+0092,Male,18,59,41
+0093,Male,48,60,49
+0094,Female,40,60,40
+0095,Female,32,60,42
+0096,Male,24,60,52
+0097,Female,47,60,47
+0098,Female,27,60,50
+0099,Male,48,61,42
+0100,Male,20,61,49
+0101,Female,23,62,41
+0102,Female,49,62,48
+0103,Male,67,62,59
+0104,Male,26,62,55
+0105,Male,49,62,56
+0106,Female,21,62,42
+0107,Female,66,63,50
+0108,Male,54,63,46
+0109,Male,68,63,43
+0110,Male,66,63,48
+0111,Male,65,63,52
+0112,Female,19,63,54
+0113,Female,38,64,42
+0114,Male,19,64,46
+0115,Female,18,65,48
+0116,Female,19,65,50
+0117,Female,63,65,43
+0118,Female,49,65,59
+0119,Female,51,67,43
+0120,Female,50,67,57
+0121,Male,27,67,56
+0122,Female,38,67,40
+0123,Female,40,69,58
+0124,Male,39,69,91
+0125,Female,23,70,29
+0126,Female,31,70,77
+0127,Male,43,71,35
+0128,Male,40,71,95
+0129,Male,59,71,11
+0130,Male,38,71,75
+0131,Male,47,71,9
+0132,Male,39,71,75
+0133,Female,25,72,34
+0134,Female,31,72,71
+0135,Male,20,73,5
+0136,Female,29,73,88
+0137,Female,44,73,7
+0138,Male,32,73,73
+0139,Male,19,74,10
+0140,Female,35,74,72
+0141,Female,57,75,5
+0142,Male,32,75,93
+0143,Female,28,76,40
+0144,Female,32,76,87
+0145,Male,25,77,12
+0146,Male,28,77,97
+0147,Male,48,77,36
+0148,Female,32,77,74
+0149,Female,34,78,22
+0150,Male,34,78,90
+0151,Male,43,78,17
+0152,Male,39,78,88
+0153,Female,44,78,20
+0154,Female,38,78,76
+0155,Female,47,78,16
+0156,Female,27,78,89
+0157,Male,37,78,1
+0158,Female,30,78,78
+0159,Male,34,78,1
+0160,Female,30,78,73
+0161,Female,56,79,35
+0162,Female,29,79,83
+0163,Male,19,81,5
+0164,Female,31,81,93
+0165,Male,50,85,26
+0166,Female,36,85,75
+0167,Male,42,86,20
+0168,Female,33,86,95
+0169,Female,36,87,27
+0170,Male,32,87,63
+0171,Male,40,87,13
+0172,Male,28,87,75
+0173,Male,36,87,10
+0174,Male,36,87,92
+0175,Female,52,88,13
+0176,Female,30,88,86
+0177,Male,58,88,15
+0178,Male,27,88,69
+0179,Male,59,93,14
+0180,Male,35,93,90
+0181,Female,37,97,32
+0182,Female,32,97,86
+0183,Male,46,98,15
+0184,Female,29,98,88
+0185,Female,41,99,39
+0186,Male,30,99,97
+0187,Female,54,101,24
+0188,Male,28,101,68
+0189,Female,41,103,17
+0190,Female,36,103,85
+0191,Female,34,103,23
+0192,Female,32,103,69
+0193,Male,33,113,8
+0194,Female,38,113,91
+0195,Female,47,120,16
+0196,Female,35,120,79
+0197,Female,45,126,28
+0198,Male,32,126,74
+0199,Male,32,137,18
+0200,Male,30,137,83

Klustering algo/hc.py

@@ -0,0 +1,49 @@
+    """
+Created on Fri Mar 31 21:41:34 2017
+
+@author: Robert
+"""
+
+import numpy as np
+import pandas as pd
+import matplotlib.pyplot as plt
+plt.style.use('seaborn-deep')
+import matplotlib.cm
+cmap = matplotlib.cm.get_cmap('plasma')
+
+# Reading in data
+ds = pd.read_csv('Mall_Customers.csv')
+X = ds.iloc[:, [3,4]].values
+
+# Dendrogram to choose number of clusters (k)
+import scipy.cluster.hierarchy as sch
+
+plt.figure(1)
+z = sch.linkage(X, method = 'ward')
+dendrogram = sch.dendrogram(z)
+plt.title('Dendrogram')
+plt.xlabel('Customers')
+plt.ylabel('Euclidean distances')
+plt.show()
+
+k = 5
+
+# Clustering
+from sklearn.cluster import AgglomerativeClustering
+
+hc = AgglomerativeClustering(n_clusters = k, affinity = "euclidean",
+                             linkage = 'ward')
+y_hc = hc.fit_predict(X)
+
+labels = [('Cluster ' + str(i+1)) for i in range(k)]
+
+plt.figure(2)
+for i in range(k):
+    plt.scatter(X[y_hc == i, 0], X[y_hc == i, 1], s = 20,
+                c = cmap(i/k), label = labels[i]) 
+plt.xlabel('Age')
+plt.ylabel('Spending score')
+plt.title('HC cluster plot')
+plt.legend()
+plt.show()
+

Klustering algo/hc.r

@@ -0,0 +1,17 @@
+
+# Reading in data
+ds = read.csv('Mall_Customers.csv')
+X = ds[,4:5]
+
+# Creating dendrogram to choose k
+hc = hclust(dist(X, method = "euclidean"), method = "ward.D")
+
+plot(hc, labels = FALSE, hang = 0.03, 
+     main = paste("Cluster Dendrogram"), 
+     xlab = 'Customers', 
+     ylab = "Euclidean distance")
+
+# Clustering
+y_hc = cutree(hc, 5)
+
+plot(X, col = y_hc)

Klustering algo/kmeans.py

@@ -0,0 +1,54 @@
+"""
+Created on Wed Mar 29 21:42:38 2017
+
+@author: Robert
+"""
+
+import numpy as np
+import pandas as pd
+import matplotlib.pyplot as plt
+plt.style.use('seaborn-deep')
+from sklearn.cluster import KMeans
+import matplotlib.cm
+cmap = matplotlib.cm.get_cmap('plasma')
+
+
+ds = pd.read_csv('Mall_Customers.csv')
+X = ds.iloc[:, [3,4]].values
+
+# Choosing the value of k by the elbow method
+wcss = []
+
+for i in range(1,21):
+    kmeans = KMeans(n_clusters=i)
+    kmeans.fit_transform(X)
+    wcss.append(kmeans.inertia_)
+
+plt.figure()
+plt.plot(range(1,21), wcss)
+plt.title('The Elbow Method')
+plt.xlabel('Number of clusters')
+plt.ylabel('WCSS')
+plt.show()
+
+# Clustering the data
+k = 5
+kmeans = KMeans(n_clusters = k)
+y_kmeans = kmeans.fit_predict(X)
+
+labels = [('Cluster ' + str(i+1)) for i in range(k)]
+
+# Plotting the clusters
+plt.figure()
+for i in range(k):
+    plt.scatter(X[y_kmeans == i, 0], X[y_kmeans == i, 1], s = 20,
+                 c = cmap(i/k), label = labels[i]) 
+
+plt.scatter(kmeans.cluster_centers_[:,0], kmeans.cluster_centers_[:,1],
+            s = 100, c = 'black', label = 'Centroids', marker = 'X')
+plt.xlabel('Age')
+plt.ylabel('Spending score')
+plt.title('Kmeans cluster plot')
+plt.legend()
+plt.show()
+

Klustering algo/kmeans.r

@@ -0,0 +1,19 @@
+
+# Reading in data
+ds = read.csv('Mall_Customers.csv')
+X = ds[4:5]
+
+# Finding k
+wcss = vector()
+for (i in 1:10) 
+    wcss[i] =sum(kmeans(X, i)$withinss)
+
+plot(1:10, wcss, type = 'b', main=paste("Elbow method"), xlab = 'number clusters' )
+
+# Clustering
+kmeans = kmeans(X, 5)
+y_kmeans = kmeans$cluster
+
+# Visualising the clusters
+plot(X, col = y_kmeans)
+points(kmeans$center,col=1:2,pch=8,cex=1)