Spark DataFrames

转自：http://www.k6k4.com/chapter/show/aafliljce1474164458328

1、样本数据

每一行存一个json对象

{    "name": "Andy",    "age": 30  }
{    "name": "Justin",    "age": 19  }
{    "name": "tom",    "age": 21  }

文件路径为 example/input/data

2、加载数据

scala> val df=spark.read.json("example/input/data")
...
df: org.apache.spark.sql.DataFrame = [age: bigint, name: string]

3、查看数据

scala> df.show
+---+------+
|age|  name|
+---+------+
| 30|  Andy|
| 19|Justin|
| 21|   tom|
+---+------+

4、查看表Schema

scala> df.printSchema
root
 |-- age: long (nullable = true)
 |-- name: string (nullable = true)

5、数据查询基本操作

scala> df.select("name").show
+------+
|  name|
+------+
|  Andy|
|Justin|
|   tom|
+------+
 
scala> df.select($"name",$"age"+1).show
+------+---------+
|  name|(age + 1)|
+------+---------+
|  Andy|       31|
|Justin|       20|
|   tom|       22|
+------+---------+
 
scala> df.filter($"age">21).show
+---+----+
|age|name|
+---+----+
| 30|Andy|
+---+----+