AggregationBuilders.terms 相当于sql中的group by

AggregationBuilders.terms("name").field("id")；

其中 name值自定义，id为需要分组的key

ValueCountAggregationBuilder countByFloor = AggregationBuilders.count("count").field("buildId.keyword");
TermsAggregationBuilder gourpByFloorId = AggregationBuilders.terms("buildId").field("buildId.keyword")
                .subAggregation(countByFloor)

根据建筑id分组，并统计每栋建筑下总共有多少个分户。可在分组是加上要统计count ，即 .subAggregation(countByFloor)；

SumBucketPipelineAggregationBuilder houseSum = PipelineAggregatorBuilders.sumBucket("houseSum", "buildId.count");

ValueCountAggregationBuilder countByFloor = AggregationBuilders.count("count").field("buildId.keyword");
TermsAggregationBuilder gourpByFloorId = AggregationBuilders.terms("buildId").field("buildId.keyword")
                .subAggregation(countByFloor);
TermsAggregationBuilder termsAggregationBuilder = AggregationBuilders.terms("cata").field("useAndClassName.keyword")
                .subAggregation(houseSum)
                .subAggregation(gourpByFloorId);                

根据buildId 聚合后，需要根据聚合后的值进行二次计算，则需要使用 PipelineAggregatorBuilders。

PipelineAggregatorBuilders.sumBucket(“houseSum”, “buildId.count”);
其中houseSum为自定义的值，buildId.count则必须是，聚合时的统计的key值