1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364"""README, Author - Anurag Kumar(mailto:anuragkumarak95@gmail.com)
Requirements:
- sklearn
- numpy
- matplotlib
Python:
- 3.5
Inputs:
- X , a 2D numpy array of features.
- k , number of clusters to create.
- initial_centroids , initial centroid values generated by utility function(mentioned
in usage).
- maxiter , maximum number of iterations to process.
- heterogeneity , empty list that will be filled with heterogeneity values if passed
to kmeans func.
Usage:
1. define 'k' value, 'X' features array and 'heterogeneity' empty list
2. create initial_centroids,
initial_centroids = get_initial_centroids(
X,
k,
seed=0 # seed value for initial centroid generation,
# None for randomness(default=None)
)
3. find centroids and clusters using kmeans function.
centroids, cluster_assignment = kmeans(
X,
k,
initial_centroids,
maxiter=400,
record_heterogeneity=heterogeneity,
verbose=True # whether to print logs in console or not.(default=False)
)
4. Plot the loss function and heterogeneity values for every iteration saved in
heterogeneity list.
plot_heterogeneity(
heterogeneity,
k
)
5. Plot the labeled 3D data points with centroids.
plot_kmeans(
X,
centroids,
cluster_assignment
)
6. Transfers Dataframe into excel format it must have feature called
'Clust' with k means clustering numbers in it.
"""
import warnings
import numpy as np
import pandas as pd
from matplotlib import pyplot as plt
from sklearn.metrics import pairwise_distances
warnings.filterwarnings("ignore")
TAG = "K-MEANS-CLUST/ "
def get_initial_centroids(data, k, seed=None):
"""Randomly choose k data points as initial centroids"""
# useful for obtaining consistent results
rng = np.random.default_rng(seed)
n = data.shape[0] # number of data points
# Pick K indices from range [0, N).
rand_indices = rng.integers(0, n, k)
# Keep centroids as dense format, as many entries will be nonzero due to averaging.
# As long as at least one document in a cluster contains a word,
# it will carry a nonzero weight in the TF-IDF vector of the centroid.
centroids = data[rand_indices, :]
return centroids
def centroid_pairwise_dist(x, centroids):
return pairwise_distances(x, centroids, metric="euclidean")
def assign_clusters(data, centroids):
# Compute distances between each data point and the set of centroids:
# Fill in the blank (RHS only)
distances_from_centroids = centroid_pairwise_dist(data, centroids)
# Compute cluster assignments for each data point:
# Fill in the blank (RHS only)
cluster_assignment = np.argmin(distances_from_centroids, axis=1)
return cluster_assignment
def revise_centroids(data, k, cluster_assignment):
new_centroids = []
for i in range(k):
# Select all data points that belong to cluster i. Fill in the blank (RHS only)
member_data_points = data[cluster_assignment == i]
# Compute the mean of the data points. Fill in the blank (RHS only)
centroid = member_data_points.mean(axis=0)
new_centroids.append(centroid)
new_centroids = np.array(new_centroids)
return new_centroids
def compute_heterogeneity(data, k, centroids, cluster_assignment):
heterogeneity = 0.0
for i in range(k):
# Select all data points that belong to cluster i. Fill in the blank (RHS only)
member_data_points = data[cluster_assignment == i, :]
if member_data_points.shape[0] > 0: # check if i-th cluster is non-empty
# Compute distances from centroid to data points (RHS only)
distances = pairwise_distances(
member_data_points, [centroids[i]], metric="euclidean"
)
squared_distances = distances**2
heterogeneity += np.sum(squared_distances)
return heterogeneity
def plot_heterogeneity(heterogeneity, k):
plt.figure(figsize=(7, 4))
plt.plot(heterogeneity, linewidth=4)
plt.xlabel("# Iterations")
plt.ylabel("Heterogeneity")
plt.title(f"Heterogeneity of clustering over time, K={k:d}")
plt.rcParams.update({"font.size": 16})
plt.show()
def plot_kmeans(data, centroids, cluster_assignment):
ax = plt.axes(projection="3d")
ax.scatter(data[:, 0], data[:, 1], data[:, 2], c=cluster_assignment, cmap="viridis")
ax.scatter(
centroids[:, 0], centroids[:, 1], centroids[:, 2], c="red", s=100, marker="x"
)
ax.set_xlabel("X")
ax.set_ylabel("Y")
ax.set_zlabel("Z")
ax.set_title("3D K-Means Clustering Visualization")
plt.show()
def kmeans(
data, k, initial_centroids, maxiter=500, record_heterogeneity=None, verbose=False
):
"""Runs k-means on given data and initial set of centroids.
maxiter: maximum number of iterations to run.(default=500)
record_heterogeneity: (optional) a list, to store the history of heterogeneity
as function of iterations
if None, do not store the history.
verbose: if True, print how many data points changed their cluster labels in
each iteration"""
centroids = initial_centroids[:]
prev_cluster_assignment = None
for itr in range(maxiter):
if verbose:
print(itr, end="")
# 1. Make cluster assignments using nearest centroids
cluster_assignment = assign_clusters(data, centroids)
# 2. Compute a new centroid for each of the k clusters, averaging all data
# points assigned to that cluster.
centroids = revise_centroids(data, k, cluster_assignment)
# Check for convergence: if none of the assignments changed, stop
if (
prev_cluster_assignment is not None
and (prev_cluster_assignment == cluster_assignment).all()
):
break
# Print number of new assignments
if prev_cluster_assignment is not None:
num_changed = np.sum(prev_cluster_assignment != cluster_assignment)
if verbose:
print(
f" {num_changed:5d} elements changed their cluster assignment."
)
# Record heterogeneity convergence metric
if record_heterogeneity is not None:
# YOUR CODE HERE
score = compute_heterogeneity(data, k, centroids, cluster_assignment)
record_heterogeneity.append(score)
prev_cluster_assignment = cluster_assignment[:]
return centroids, cluster_assignment
# Mock test below
if False: # change to true to run this test case.
from sklearn import datasets as ds
dataset = ds.load_iris()
k = 3
heterogeneity = []
initial_centroids = get_initial_centroids(dataset["data"], k, seed=0)
centroids, cluster_assignment = kmeans(
dataset["data"],
k,
initial_centroids,
maxiter=400,
record_heterogeneity=heterogeneity,
verbose=True,
)
plot_heterogeneity(heterogeneity, k)
plot_kmeans(dataset["data"], centroids, cluster_assignment)
def report_generator(
predicted: pd.DataFrame, clustering_variables: np.ndarray, fill_missing_report=None
) -> pd.DataFrame:
"""
Generate a clustering report given these two arguments:
predicted - dataframe with predicted cluster column
fill_missing_report - dictionary of rules on how we are going to fill in missing
values for final generated report (not included in modelling);
>>> predicted = pd.DataFrame()
>>> predicted['numbers'] = [1, 2, 3]
>>> predicted['col1'] = [0.5, 2.5, 4.5]
>>> predicted['col2'] = [100, 200, 300]
>>> predicted['col3'] = [10, 20, 30]
>>> predicted['Cluster'] = [1, 1, 2]
>>> report_generator(predicted, ['col1', 'col2'], 0)
Features Type Mark 1 2
0 # of Customers ClusterSize False 2.000000 1.000000
1 % of Customers ClusterProportion False 0.666667 0.333333
2 col1 mean_with_zeros True 1.500000 4.500000
3 col2 mean_with_zeros True 150.000000 300.000000
4 numbers mean_with_zeros False 1.500000 3.000000
.. ... ... ... ... ...
99 dummy 5% False 1.000000 1.000000
100 dummy 95% False 1.000000 1.000000
101 dummy stdev False 0.000000 NaN
102 dummy mode False 1.000000 1.000000
103 dummy median False 1.000000 1.000000
<BLANKLINE>
[104 rows x 5 columns]
"""
# Fill missing values with given rules
if fill_missing_report:
predicted = predicted.fillna(value=fill_missing_report)
predicted["dummy"] = 1
numeric_cols = predicted.select_dtypes(np.number).columns
report = (
predicted.groupby(["Cluster"])[ # construct report dataframe
numeric_cols
] # group by cluster number
.agg(
[
("sum", "sum"),
("mean_with_zeros", lambda x: np.mean(np.nan_to_num(x))),
("mean_without_zeros", lambda x: x.replace(0, np.nan).mean()),
(
"mean_25-75",
lambda x: np.mean(
np.nan_to_num(
sorted(x)[
round(len(x) * 25 / 100) : round(len(x) * 75 / 100)
]
)
),
),
("mean_with_na", "mean"),
("min", lambda x: x.min()),
("5%", lambda x: x.quantile(0.05)),
("25%", lambda x: x.quantile(0.25)),
("50%", lambda x: x.quantile(0.50)),
("75%", lambda x: x.quantile(0.75)),
("95%", lambda x: x.quantile(0.95)),
("max", lambda x: x.max()),
("count", lambda x: x.count()),
("stdev", lambda x: x.std()),
("mode", lambda x: x.mode()[0]),
("median", lambda x: x.median()),
("# > 0", lambda x: (x > 0).sum()),
]
)
.T.reset_index()
.rename(index=str, columns={"level_0": "Features", "level_1": "Type"})
) # rename columns
# calculate the size of cluster(count of clientID's)
# avoid SettingWithCopyWarning
clustersize = report[
(report["Features"] == "dummy") & (report["Type"] == "count")
].copy()
# rename created predicted cluster to match report column names
clustersize.Type = "ClusterSize"
clustersize.Features = "# of Customers"
# calculating the proportion of cluster
clusterproportion = pd.DataFrame(
clustersize.iloc[:, 2:].to_numpy() / clustersize.iloc[:, 2:].to_numpy().sum()
)
# rename created predicted cluster to match report column names
clusterproportion["Type"] = "% of Customers"
clusterproportion["Features"] = "ClusterProportion"
cols = clusterproportion.columns.tolist()
cols = cols[-2:] + cols[:-2]
clusterproportion = clusterproportion[cols] # rearrange columns to match report
clusterproportion.columns = report.columns
# generating dataframe with count of nan values
a = pd.DataFrame(
abs(
report[report["Type"] == "count"].iloc[:, 2:].to_numpy()
- clustersize.iloc[:, 2:].to_numpy()
)
)
a["Features"] = 0
a["Type"] = "# of nan"
# filling values in order to match report
a.Features = report[report["Type"] == "count"].Features.tolist()
cols = a.columns.tolist()
cols = cols[-2:] + cols[:-2]
a = a[cols] # rearrange columns to match report
a.columns = report.columns # rename columns to match report
# drop count values except for cluster size
report = report.drop(report[report.Type == "count"].index)
# concat report with cluster size and nan values
report = pd.concat([report, a, clustersize, clusterproportion], axis=0)
report["Mark"] = report["Features"].isin(clustering_variables)
cols = report.columns.tolist()
cols = cols[0:2] + cols[-1:] + cols[2:-1]
report = report[cols]
sorter1 = {
"ClusterSize": 9,
"ClusterProportion": 8,
"mean_with_zeros": 7,
"mean_with_na": 6,
"max": 5,
"50%": 4,
"min": 3,
"25%": 2,
"75%": 1,
"# of nan": 0,
"# > 0": -1,
"sum_with_na": -2,
}
report = (
report.assign(
Sorter1=lambda x: x.Type.map(sorter1),
Sorter2=lambda x: list(reversed(range(len(x)))),
)
.sort_values(["Sorter1", "Mark", "Sorter2"], ascending=False)
.drop(["Sorter1", "Sorter2"], axis=1)
)
report.columns.name = ""
report = report.reset_index()
report = report.drop(columns=["index"])
return report
if __name__ == "__main__":
import doctest
doctest.testmod()