Clustering example using cities

c go java js lua py vbs rkt scm pro

Here is an example of clustering using the geographic coordinates of some cities and towns in Pennsylvania.

For the data transformation, see Location text data conversion .

Here is the Python code.

import os
import matplotlib as mpl
import matplotlib.pyplot as plt
from PIL import Image

import numpy as np
import sklearn as sk
from sklearn.datasets import make_blobs
from sklearn.cluster import KMeans

import json

imagePos1 = 0
rlen1 = 1
clen1 = 1
dpi1 = 72
width1 = 640
height1 = 480
figSize1 = (width1 / dpi1, height1/dpi1)
fig1, ax1 = plt.subplots(rlen1,clen1,figsize=figSize1,dpi=dpi1)

fig1,ax1 = plt.subplots(1,1)

centers1 = 6
jsonFs1 = "././mapdata1.json"
print("Read: \"{0:s}\"".format(jsonFs1))
with open(jsonFs1,"r") as jsonF1:
	jsonDict1 = json.load(jsonF1)

dataList1 = jsonDict1["data"]
dataList2 = []
for (lat1,lng1,city1) in dataList1:
	dataList2.append([lng1,lat1])
dataArray2 = np.array(dataList2)
kmeans1 = KMeans(
	n_clusters=centers1,
	init="k-means++",
	max_iter=300,
	n_init=10,
	random_state=0,
	)
pred_y = kmeans1.fit_predict(dataArray2)

plt.scatter(
	dataArray2[:,0],
	dataArray2[:,1],
	c="green",
	)

plt.scatter(
	kmeans1.cluster_centers_[:, 0],
	kmeans1.cluster_centers_[:, 1],
	s=200,
	c="red",
	)

plt.xlabel("Latitude")
plt.ylabel("Longitude")
plt.title("Cities scatter plot")

fig1.tight_layout()
imagePos1 += 1
type1 = "png"
saveFs1 = "./group23-1-0{0:d}.{1:s}".format(imagePos1,type1)
fig1.savefig(saveFs1,bbox_inches="tight")
savedImage1 = Image.open(saveFs1)
savedWidth1, savedHeight1= savedImage1.size
(_, _, _, _, _, _, savedSize1, _, savedTime1, _) = os.stat(saveFs1)
print("\t{0:s} : {1:d} x {2:d} , {3:,} bytes".format(saveFs1,savedWidth1,savedHeight1,savedSize1))

Here is the output of the Python code.

Read: "././mapdata1.json"
	./group23-1-01.png : 630 x 470 , 66,496 bytes