如何在impala或hive上读取复杂类型的数组<string>?

时间:2018-02-23 13:11:59

标签: hadoop hive parquet impala

我尝试在hive上阅读复杂类型

eval

使用以下查询

dog.owners (array<string>)

但我收到以下错误

  

select dog_id, concat_ws(',',collect_set(owners)) as owners from dog group by dog_id

看起来数据类型不匹配。我尝试将列所有者创建为Argument 2 of function CONCAT_WS must be "string or array<string>", but "array<array<string>>" was found.,但我仍然遇到相同的错误。 有没有办法在蜂巢或黑斑羚上阅读该专栏?

1 个答案:

答案 0 :(得分:1)

select dog_id, do.* from dog, dog.owners as do