spark 决策树分类算法demo

分类（Classification）

下面的例子说明了怎样导入LIBSVM 数据文件，解析成RDD[LabeledPoint]，然后使用决策树进行分类。GINI不纯度作为不纯度衡量标准并且树的最大深度设置为5。最后计算了测试错误率从而评估算法的准确性。

from pyspark.mllib.regression import LabeledPoint
from pyspark.mllib.tree import DecisionTree, DecisionTreeModel
from pyspark.mllib.util import MLUtils

# Load and parse the data file into an RDD of LabeledPoint.
data = MLUtils.loadLibSVMFile(sc, 'data/mllib/sample_libsvm_data.txt')
# Split the data into training and test sets (30% held out for testing)
(trainingData, testData) = data.randomSplit([0.7, 0.3])

# Train a DecisionTree model.
#  Empty categoricalFeaturesInfo indicates all features are continuous.
model = DecisionTree.trainClassifier(trainingData, numClasses=2, categoricalFeaturesInfo={},
                                     impurity='gini', maxDepth=5, maxBins=32)

# Evaluate model on test instances and compute test error
predictions = model.predict(testData.map(lambda x: x.features))
labelsAndPredictions = testData.map(lambda lp: lp.label).zip(predictions)
testErr = labelsAndPredictions.filter(lambda (v, p): v != p).count() / float(testData.count())
print('Test Error = ' + str(testErr))
print('Learned classification tree model:')
print(model.toDebugString())

# Save and load model
model.save(sc, "myModelPath")
sameModel = DecisionTreeModel.load(sc, "myModelPath")

以下代码展示了如何载入一个LIBSVM数据文件，解析成一个LabeledPointRDD，然后使用决策树，使用Gini不纯度作为不纯度衡量指标，最大树深度是5.测试误差用来计算算法准确率。

1. # -*- coding:utf-8 -*-
2. """
3. 测试决策树
4. """
6. import os
7. import sys
8. import logging
9. from pyspark.mllib.tree import DecisionTree,DecisionTreeModel
10. from pyspark.mllib.util import MLUtils
12. # Path for spark source folder
13. os.environ['SPARK_HOME']="D:\javaPackages\spark-1.6.0-bin-hadoop2.6"
14. # Append pyspark to Python Path
15. sys.path.append("D:\javaPackages\spark-1.6.0-bin-hadoop2.6\python")
16. sys.path.append("D:\javaPackages\spark-1.6.0-bin-hadoop2.6\python\lib\py4j-0.9-src.zip")
18. from pyspark import SparkContext
19. from pyspark import SparkConf
21. conf = SparkConf()
22. conf.set("YARN_CONF_DIR ", "D:\javaPackages\hadoop_conf_dir\yarn-conf")
23. conf.set("spark.driver.memory", "2g")
25. #conf.set("spark.executor.memory", "1g")
26. #conf.set("spark.python.worker.memory", "1g")
27. conf.setMaster("yarn-client")
28. conf.setAppName("TestDecisionTree")
29. logger = logging.getLogger('pyspark')
30. sc = SparkContext(conf=conf)
32. mylog = []
33. #载入和解析数据文件为 LabeledPoint RDDdata = MLUtils.loadLibSVMFile(sc,"/home/xiatao/machine_learing/")
34. #将数据拆分成训练集合测试集
35. (trainingData,testData) = data.randomSplit([0.7,0.3])
37. ##训练决策树模型
38. #空的 categoricalFeauresInfo 代表了所有的特征都是连续的
39. model = DecisionTree.trainClassifier(trainingData, numClasses=2,categoricalFeaturesInfo={},impurity='gini',maxDepth=5,maxBins=32)
41. # 在测试实例上评估模型并计算测试误差
43. predictions = model.predict(testData.map(lambda x:x.features))
44. labelsAndPoint = testData.map(lambda lp:lp.label).zip(predictions)
45. testMSE = labelsAndPoint.map(lambda (v,p):(v-p)**2).sum()/float(testData.count())
46. mylog.append("测试误差是")
47. mylog.append(testMSE)
49. #存储模型
51. model.save(sc,"/home/xiatao/machine_learing/")
52. sc.parallelize(mylog).saveAsTextFile("/home/xiatao/machine_learing/log")
53. sameModel = DecisionTreeModel.load(sc,"/home/xiatao/machine_learing/")