一、通过结构化数据创建DataFrame：
publicstaticvoid main(String[] args) {
SparkConf conf = new SparkConf()
.setAppName("DataFrameCreate").setMaster("local");
JavaSparkContext sc = new JavaSparkContext(conf);
SQLContext sqlContext = new SQLContext(sc);
DataFrame df = sqlContext.read().json("hdfs://spark1:9000/students.json"); //结构化数据直接加载为DataFrame
df.show();
}
二、通过RDD创建DataFrame的两种创建方式
（数据源students.txt的数据截图）
2.1通过已知类型的schema创建DataFrame，代码如下：
public static void main(String[] args) {
SparkConf conf = new SparkConf()
.setMaster("local")
.setAppName("RDD2DataFrameReflection");
JavaSparkContext sc = new JavaSparkContext(conf);
SQLContext sqlContext = new SQLContext(sc);
JavaRDD<String> lines = sc.textFile("D://students.txt");
//将lines转换成 JavaRDD<Student>
JavaRDD<Student> students = lines.map(new Function<String, Student>() {
private static final long serialVersionUID = 1L;
@Override
public Student call(String line) throws Exception {
// TODO Auto-generated method stub
String[] strPlits = line.split(",");
Student stu = new Student();
stu.setId(Integer.valueOf(strPlits[0]));
stu.setName(strPlits[1]);
stu.setAge(Integer.valueOf(strPlits[2]));
return stu;
}
});
// 使用反射方式，将RDD转换为DataFrame
// 这里要求，JavaBean必须实现Serializable接口，是可序列化的
//根据student的schema 和 RDD创建DataFrame
DataFrame studentsDF = sqlContext.createDataFrame(students, Student.class);
studentsDF.show();
}
2.2手动创建schema的方式创建DataFrame
public static void main(String[] args) {
//... 省略创建sqlContext的过程
JavaRDD<String> lines = sc.textFile("D://students.txt");
//将普通RDD装换成JavaRDD<Row>
JavaRDD<Row> rowRDD = lines.map(new Function<String, Row>() {
private static final long serialVersionUID = 1L;
@Override
public Row call(String line) throws Exception {
String[] strArray = line.split(",");
Row row= RowFactory.create(
Integer.valueOf(strArray[0]), //id
strArray[1], //name
Integer.valueOf(strArray[2])); //age
return row;
}
});
//第二步 创建元类型， 即创建schema
List<StructField> structFields = new ArrayList<StructField>();
structFields.add(DataTypes.createStructField("id", DataTypes.IntegerType, true));
structFields.add(DataTypes.createStructField("name", DataTypes.StringType, true));
structFields.add(DataTypes.createStructField("age", DataTypes.IntegerType, true));
StructType structType = DataTypes.createStructType(structFields);
//根据元数据类型将JavaRDD<Row>转化成DataFrame
DataFrame studentDF = sqlCotnext.createDataFrame(rowRDD, structType);
studentDF.show();
}
-》DataFrame、RDD、List互转
JavaRDD<Row> rows = studentDF.javaRDD();
List<Row> studentList = rows.collect();
三、DataFrame基本用法
// 打印DataFrame中所有的数据（select * from ...）
df.show();
// 打印DataFrame的元数据（Schema）
df.printSchema();
// 查询某列所有的数据
df.select("name").show();
// 查询某几列所有的数据，并对列进行计算
df.select(df.col("name"), df.col("age").plus(1)).show();
// 根据某一列的值进行过滤
df.filter(df.col("age").gt(18)).show();
// 根据某一列进行分组，然后进行聚合
df.groupBy(df.col("age")).count().show();
DataFrame studentDF = sqlCotnext.createDataFrame(rowRDD, structType);
studentDF.show();
studentDF.registerTempTable("students"); //将DataFrame注册为零时表，取名students
//对students零时表做sql查询
DataFrame oldStudentDF = sqlCotnext.sql("select * from students where age>18");
oldStudentDF.show();