我正在使用scala对spark进行一些测试.我们通常会读取需要操作的json文件,如下例所示:
test.json:
{"a":1,"b":[2,3]}
Run Code Online (Sandbox Code Playgroud)
val test = sqlContext.read.json("test.json")
Run Code Online (Sandbox Code Playgroud)
如何将其转换为以下格式:
{"a":1,"b":2}
{"a":1,"b":3}
Run Code Online (Sandbox Code Playgroud) 我有一个具有以下结构的数据帧:
|-- data: struct (nullable = true)
| |-- id: long (nullable = true)
| |-- keyNote: struct (nullable = true)
| | |-- key: string (nullable = true)
| | |-- note: string (nullable = true)
| |-- details: map (nullable = true)
| | |-- key: string
| | |-- value: string (valueContainsNull = true)
Run Code Online (Sandbox Code Playgroud)
如何展平结构并创建新的数据框:
|-- id: long (nullable = true)
|-- keyNote: struct (nullable = true)
| |-- key: string (nullable = true)
| |-- note: …Run Code Online (Sandbox Code Playgroud)