Apache hive表的列定义如下:
myvars:array<struct<index:bigint,value:string>>
对应数据的示例如下:
"myvars":[
{"index":2,"value":"value1"}
, {"index":1,"value":"value2"}
, {"index":2,"value":"value3"}
]
这个数组如何被筛选到所有"index"==2的元素
在JavaScript中,我将做如下操作:myvars.filter(function(d){return d.index==2;})
如何实现相同的结果与Apache Hive QL,最好没有横向视图?
在hive中有一组Collection函数:
Collection
array_contains(Array<T> a, val)
array<K.V> map_keys(Map<K.V> a)
array<K.V> map_values(Map<K.V> a)
size(Map<K.V>|Array<T> a)
sort_array(Array<T> a)
在查询中使用
...
WHERE
array_contains(myvars,2)
我想如果你想提取索引为2的所有值,你想要这样的东西:
SELECT DISTINCT value
FROM mytable
LATERAL VIEW EXPLODE(myvars) exploded_myvars AS idx, value
WHERE idx = 2;
如果数据类型为array<map<string,string>>
,则为
SELECT DISTINCT mv["value"]
FROM mytable
LATERAL VIEW EXPLODE(myvars) exploded_myvars AS mv
WHERE mv["index"] = 2;