Skip to content

[Feature] 向量化与非向量化不一致的case汇总 #294

Open

Description

@BiteTheDDDDt

Search before asking

  • I had searched in the issues and found no similar issues.

Description

  • 向一些计算函数(如sum、float)在一些情况下浮点数精度会与非向量化有微小的误差.
  • 存储层非向量化会把null作为最小值参与聚合函数,如含有null的min列最终输出结果会是null. 向量化的Null元素则不会参与聚合计算.(似乎不能稳定复现,有待观察) -> 行存的Bug [Bug] The null value of the min function of the storage layer is incorrectly set  apache/doris#7436
  • 时间类型溢出的处理 [Bug] Inconsistent results on DATE_ADD over limit  apache/doris#6581
  • bitmap/hll 在存储层的聚合:在向量化中如果输入均为null,则输出null,在行存中,如果输入均为null,返回空集
  • varchar和decimal计算的逻辑 非向量化目前会走varchar->double->decimal 而varchar转decimal的parser和转double的逻辑不一样 可能会有结果diff 特别是一些parse fail的情况
  • 空的Bitmap求bitmap_min会返回0,行存则会返回null.
  • Date/DateTime类型to string时时分秒的显示问题,需要基于function context重写这个逻辑,或者彻底重构这两个类型。

Use case

No response

Related issues

No response

Are you willing to submit PR?

  • Yes I am willing to submit a PR!

Code of Conduct

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions