Merge pull request #229 from jiangzhonglian/master

更新 数字识别  knn 写入方式

competitions/getting-started/house-price/README.md

@@ -1929,6 +1929,10 @@ mode_br.fit(x_train, y_train)
 y_test = np.expm1(mode_br.predict(x_test))
 ```
 
+## 四 建立模型
+
+> 模型融合 voting
+
 ```python
 # 模型融合
 class AveragingModels(BaseEstimator, RegressorMixin, TransformerMixin):
@@ -1989,15 +1993,3 @@ result['SalePrice'] = ensemble
 # index=False 是用来除去行编号
 result.to_csv('/Users/liudong/Desktop/house_price/result.csv', index=False)
 ```
-
-        Id      SalePrice
-    0  1461  110469.586157
-    1  1462  148368.953437
-    2  1463  172697.673678
-    3  1464  189844.587562
-    4  1465  207009.716532
-    5  1466  188820.407208
-    6  1467  163107.556014
-    7  1468  180732.346459
-    8  1469  194841.804925
-    9  1470  110570.281362

src/python/getting-started/digit-recognizer/knn-python3.6.py

@@ -14,9 +14,9 @@ import numpy as np
 import pandas as pd
 from sklearn.decomposition import PCA
 from sklearn.neighbors import KNeighborsClassifier
-import sys
 
-data_dir = '/Users/wuyanxue/Documents/GitHub/datasets/getting-started/digit-recognizer/'
+data_dir = '/opt/data/kaggle/getting-started/digit-recognizer/'
+
 
 # 加载数据
 def opencsv():
@@ -31,18 +31,15 @@ def opencsv():
 
 
 def saveResult(result, csvName):
-    with open(csvName, 'w', newline='') as myFile:  # 创建记录输出结果的文件（w 和 wb 使用的时候有问题）
+    with open(csvName, 'w') as myFile:  # 创建记录输出结果的文件（w 和 wb 使用的时候有问题）
         # python3里面对 str和bytes类型做了严格的区分，不像python2里面某些函数里可以混用。所以用python3来写wirterow时，打开文件不要用wb模式，只需要使用w模式，然后带上newline=''
-        myWriter = csv.writer(myFile)  # 对文件执行写入
+        myWriter = csv.writer(myFile)
-        myWriter.writerow(["ImageId", "Label"])  # 设置表格的列名
+        myWriter.writerow(["ImageId", "Label"])
         index = 0
-        for i in result:
+        for r in result:
-            tmp = []
+            index += 1
-            index = index + 1
+            myWriter.writerow([index, int(r)])
-            tmp.append(index)
+    print('Saved successfully...')  # 保存预测结果
-            # tmp.append(i)
-            tmp.append(int(i))  # 测试集的标签值
-            myWriter.writerow(tmp)
 
 
 def knnClassify(trainData, trainLabel):
@@ -95,7 +92,7 @@ def dRecognition_knn():
 
     # 结果预测
     testLabel = knnClf.predict(testData)
-    
+
     # 结果的输出
     saveResult(testLabel, os.path.join(data_dir, 'output/Result_sklearn_knn.csv'))
     print("finish!")

src/python/getting-started/digit-recognizer/svm-python3.6.py

@@ -6,7 +6,6 @@ Created on 2017-10-26
 Update  on 2017-10-26
 Author: 片刻
 Github: https://github.com/apachecn/kaggle
-PCA主成成分分析
 '''
 
 import os.path
@@ -21,7 +20,8 @@ from sklearn.metrics import classification_report
 from sklearn.model_selection import train_test_split
 
 # 数据路径
-data_dir = '/Users/wuyanxue/Documents/GitHub/datasets/getting-started/digit-recognizer/'
+data_dir = '/opt/data/kaggle/getting-started/digit-recognizer/'
+
 
 # 加载数据
 def opencsv():
@@ -61,7 +61,6 @@ def dRCsv(x_train, x_test, preData, COMPONENT_NUM):
     return pcaTrainData,  pcaTestData, pcaPreData
 
 
-
 # 训练模型
 def trainModel(trainData, trainLabel):
     print('Train SVM...')
@@ -85,20 +84,20 @@ def saveResult(result, csvName):
 # 分析数据,看数据是否满足要求（通过这些来检测数据的相关性，考虑在分类的时候提取出重要的特征）
 def analyse_data(dataMat):
     meanVals = np.mean(dataMat, axis=0)  # np.mean 求出每列的平均值meanVals
-    meanRemoved = dataMat-meanVals # 每一列特征值减去该列的特征值均值
+    meanRemoved = dataMat-meanVals  # 每一列特征值减去该列的特征值均值
     # 计算协方差矩阵，除数n-1是为了得到协方差的 无偏估计
     # cov(X,0) = cov(X) 除数是n-1(n为样本个数)
     # cov(X,1) 除数是n
-    covMat = np.cov(meanRemoved, rowvar=0) # cov 计算协方差的值,
+    covMat = np.cov(meanRemoved, rowvar=0)  # cov 计算协方差的值,
     # np.mat 是用来生成一个矩阵的
     # 保存特征值(eigvals)和对应的特征向量(eigVects)
-    eigvals, eigVects = np.linalg.eig(np.mat(covMat)) # linalg.eig 计算的值是矩阵的特征值，保存在对应的矩阵中
+    eigvals, eigVects = np.linalg.eig(np.mat(covMat))  # linalg.eig 计算的值是矩阵的特征值，保存在对应的矩阵中
-    eigValInd = np.argsort(eigvals) #  argsort 对特征值进行排序，返回的是数值从小到大的索引值
+    eigValInd = np.argsort(eigvals)  # argsort 对特征值进行排序，返回的是数值从小到大的索引值
 
-    topNfeat = 100 # 需要保留的特征维度，即要压缩成的维度数
+    topNfeat = 100  # 需要保留的特征维度，即要压缩成的维度数
 
     # 从排序后的矩阵最后一个开始自下而上选取最大的N个特征值，返回其对应的索引
-    eigValInd = eigValInd[:-(topNfeat+1):-1] 
+    eigValInd = eigValInd[:-(topNfeat+1):-1]
 
     # 计算特征值的总和
     cov_all_score = float(sum(eigvals))
@@ -184,6 +183,7 @@ def getModel(filename):
     fr = open(filename, 'rb')
     return pickle.load(fr)
 
+
 def trainDRSVM():
     startTime = time.time()
 
@@ -215,6 +215,7 @@ def preDRSVM():
     stopTime = time.time()
     print('PreModel load time used:%f s' % (stopTime - startTime))
 
+
 # 数据可视化
 def dataVisulization(data, labels):
     pca = PCA(n_components=2, whiten=True) # 使用PCA方法降到2维
@@ -230,6 +231,7 @@ def dataVisulization(data, labels):
     plt.title('MNIST visualization')
     plt.show()
 
+
 if __name__ == '__main__':
     trainData, trainLabel, preData = opencsv()
     dataVisulization(trainData, trainLabel)