当我使用 deeplearning4j 并尝试在 Spark 中训练模型时
public MultiLayerNetwork fit(JavaRDD<DataSet> trainingData)
fit() 需要一个 JavaRDD 参数,我尝试像这样构建
val totalDaset = csv.map(row => {
val features = Array(
row.getAs[String](0).toDouble, row.getAs[String](1).toDouble
)
val labels = Array(row.getAs[String](21).toDouble)
val featuresINDA = Nd4j.create(features)
val labelsINDA = Nd4j.create(labels)
new DataSet(featuresINDA, labelsINDA)
})
但是 IDEA 的提示是No implicit arguments of type:Encode[DataSet]
这是一个错误,我不知道如何解决这个问题,
我知道 SparkRDD 可以转换为 JavaRDD,但我不知道如何构建 Spark RDD[DataSet]
DataSet 在import org.nd4j.linalg.dataset.DataSet
它的构造方法是
public DataSet(INDArray first, INDArray second) {
this(first, second, (INDArray)null, (INDArray)null);
}
这是我的代码
val spark:SparkSession = {SparkSession
.builder()
.master("local")
.appName("Spark LSTM Emotion Analysis")
.getOrCreate()
}
import spark.implicits._
val JavaSC = JavaSparkContext.fromSparkContext(spark.sparkContext)
val csv=spark.read.format("csv")
.option("header","true")
.option("sep",",")
.load("/home/hadoop/sparkjobs/LReg/data.csv")
val totalDataset = csv.map(row => {
val features = Array(
row.getAs[String](0).toDouble, row.getAs[String](1).toDouble
)
val labels = Array(row.getAs[String](21).toDouble)
val featuresINDA = Nd4j.create(features)
val labelsINDA = Nd4j.create(labels)
new DataSet(featuresINDA, labelsINDA)
})
val data = totalDataset.toJavaRDD
在 deeplearning4j 官方指南中通过 Java 创建 JavaRDD[DataSet]:
String filePath = "hdfs:///your/path/some_csv_file.csv";
JavaSparkContext sc = new JavaSparkContext();
JavaRDD<String> rddString = sc.textFile(filePath);
RecordReader recordReader = new CSVRecordReader(',');
JavaRDD<List<Writable>> rddWritables = rddString.map(new StringToWritablesFunction(recordReader));
int labelIndex = 5; //Labels: a single integer representing the class index in column number 5
int numLabelClasses = 10; //10 classes for the label
JavaRDD<DataSet> rddDataSetClassification = rddWritables.map(new DataVecDataSetFunction(labelIndex, numLabelClasses, false));
我尝试通过 scala 创建:
val JavaSC: JavaSparkContext = new JavaSparkContext()
val rddString: JavaRDD[String] = JavaSC.textFile("/home/hadoop/sparkjobs/LReg/hf-data.csv")
val recordReader: CSVRecordReader = new CSVRecordReader(',')
val rddWritables: JavaRDD[List[Writable]] = rddString.map(new StringToWritablesFunction(recordReader))
val featureColnum = 3
val labelColnum = 1
val d = new DataVecDataSetFunction(featureColnum,labelColnum,true,null,null)
// val rddDataSet: JavaRDD[DataSet] = rddWritables.map(new DataVecDataSetFunction(featureColnum,labelColnum, true,null,null))
// can not reslove overloaded method 'map'
调试错误信息: