python kNN算法实现MNIST数据集分类 k值1-120

# -*- coding: utf-8 -*- """ Input: inX: vector to compare to existing dataset (1xN) dataSet: size m data set of known vectors (NxM) labels: data set labels (1xM vector) k: number of neighbors to use for comparison Output: the most popular class label """ from numpy import * import operator import os from data_util import DataUtils import datetime import time import matplotlib.pyplot as plt import numpy as np trainfile_X = 'train-images.idx3-ubyte' trainfile_y = 'train-labels.idx1-ubyte' testfile_X = 't10k-images.idx3-ubyte' testfile_y = 't10k-labels.idx1-ubyte' # 定义kNN分类函数 def kNNClassify(newInput, dataSet, labels, k): numSamples = dataSet.shape[0] # shape[0] 代表行数 ## step 1: 计算欧式距离 # tile(A, reps): Construct an array by repeating A reps times # the following copy numSamples rows for dataSet diff = tile(newInput, (numSamples, 1)) - dataSet # Subtract element-wise squaredDiff = diff ** 2 # squared for the subtract squaredDist = sum(squaredDiff, axis = 1) # sum is performed by row distance = squaredDist ** 0.5 ## step 2: 计算距离 # argsort() returns the indices that would sort an array in a ascending order sortedDistIndices = argsort(distance) classCount = {} # 定义一个字典用于放入元素 for i in xrange(k): ## step 3: 选择k值对应的距离 voteLabel = labels[sortedDistIndices[i]] ## step 4: 记录标签出现的次数 # when the key voteLabel is not in dictionary classCount, get() # will return 0 classCount[voteLabel] = classCount.get(voteLabel, 0) + 1 ## step 5: 返回对应标签最多的类别 maxCount = 0 for key, value in classCount.items(): if value > maxCount: maxCount = value maxIndex = key return maxIndex # 将图片转化为向量形式 def img2vector(filename): rows = 32 cols = 32 imgVector = zeros((1, rows * cols)) fileIn = open(filename) for row in xrange(rows): lineStr = fileIn.readline() for col in xrange(cols): imgVector[0, row * 32 + col] = int(lineStr[col]) return imgVector # 测试HandWriting类 def testHandWritingClass(): ## step 1: 加载数据 ISOTIMEFORMAT='%Y-%m-%d %X' load_time=time.strftime(ISOTIMEFORMAT, time.localtime()) print "step 1: loading data...", print"load_time:%s"%load_time train_x=DataUtils(filename=trainfile_X).getImage() train_y=DataUtils(filename=trainfile_y).getLabel() test_x=DataUtils(testfile_X).getImage() test_y=DataUtils(testfile_y).getLabel() ## step 2: 训练 train_time=time.strftime(ISOTIMEFORMAT, time.localtime()) print "step 2: training...", print "train_time:%s"%train_time pass ## step 3: 测试 test_time=time.strftime(ISOTIMEFORMAT, time.localtime()) print "step 3: testing...", print "test_time:%s"%test_time numTestSamples = test_x.shape[0] matchCount = 0 result=[] for m in range(1,121): for i in xrange(numTestSamples): predict = kNNClassify(test_x[i], train_x, train_y, m) if predict == test_y[i]: matchCount += 1 error = 1 - float(matchCount) / numTestSamples matchCount=0 result_time=time.strftime(ISOTIMEFORMAT, time.localtime()) print m,"result_time:%s"%result_time print 'The classify error is: %.4f' %error result.append(error) ##step4:出图 x = range(1,121) y = result plt.plot(x, y) plt.show() if __name__ == '__main__': testHandWritingClass()