我使用Python將CSV轉換爲LIBSVM數據格式。 LIBSVM的格式如下所示。第一列是目標。帶LIBSVM數據錯誤的Spark決策樹
0 0:1 1:2 2:1 4:11 6:4 7:7 8:1 9:99 10:70 11:1
0 0:1 1:2 2:1 4:8 5:1 6:3 7:7 8:1 9:99 10:62 11:1
我在MLLib決策樹代碼在Spark是
from pyspark.mllib.tree import DecisionTree, DecisionTreeModel
from pyspark.mllib.util import MLUtils
from pyspark import SparkContext
sc = SparkContext()
data = MLUtils.loadLibSVMFile(sc,"/folder/libdata.txt")
(trainingData, testData) = data.randomSplit([0.7, 0.3])
model = DecisionTree.trainClassifier(trainingData)
print(model.toDebugString())
model.save(sc, "/folder/myDecisionTreeClassificationModel")
我得到的錯誤是
java.lang.IllegalArgumentException: requirement failed: You provided 12 indices and values, which exceeds the specified vector size 11
不知道它是怎麼了。數據的格式也是正確的。
謝謝!