My K Nearest Neighbor Algorithm...without using SKLearn...
I created a KNN Algorithm by making my own data set with tuples storing the x-value, y-value, and color of 9 points, asking the user for the x and y-values of their data point and how many dots are being taken into account, then determining the distance between the point and the ones in the data set. I stored these distances in a dictionary, with the points being the keys and the distances being the values, and sorted it from the lowest distance to the highest. After, I totaled up the number of points with each color out of the inputted number of nearest points. Finally, I classified the inputted point based on which of the colors was the most prominent out of those points.
#Importing
import numpy as np
import math as math
import operator
#Data
data = [(1, 2, "red"), (10, 8, "blue"), (5, 6, "green"), (4, 1, "red"), (8, 8, "blue"), (3, 1, "green"), (1, 2, "red"), (8, 10, "blue"), (6, 5, "green")]
#Store Points(X/Y Values)
x = int(input("What's your x-value?"))
y = int(input("What's your y-value?"))
#Store # of points being taken into account(K Value)
k= int(input("How many closest dots do you want to take into account?"))
point = (x, y)
print(point)
#Distance Formula
def distance(x2, y2):
return math.sqrt((x2-x)**2+(y2-y)**2)
#Find/Store Distances of All Points(Loop/Dictionary)
knndict = {}
for n in data:
z = distance(n[0],n[1])
print(z)
knndict[n] = z
#Ordering The Dictionary(Least Distance)
knndicts = sorted(knndict.items(), key=operator.itemgetter(1))
print(knndicts)
#Voting Process
green = 0
blue = 0
red = 0
for n in range(0, k, 1):
if knndicts[n][0][2] == "green":
green += 1
if knndicts[n][0][2] == "blue":
blue += 1
if knndicts[n][0][2] == "red":
red += 1
print(green)
print(blue)
print(red)
#Determining the Color
if green > blue and green > red:
print("Your dot's color has been classified as green!")
elif blue > green and blue > red:
print("Your dot's color has been classified as blue!")
elif red > green and red > blue:
print("Your dot's color has been classified as red!")
else:
print("Unfortunately, your dot could not be classified. Try picking a different number of closest dots next time.")