[Enhancement](doris-future) Support "REGR_" aggregation functions (PART II)

[Yoruet]

Proposed changes

Issue Number: close #38975

mysql> select * from test;
+------+------+------+
| id   | x    | y    |
+------+------+------+
|    1 |   18 |   13 |
|    3 |   12 |    2 |
|    5 |   10 |   20 |
|    2 |   14 |   27 |
|    4 |    5 |    6 |
+------+------+------+
5 rows in set (0.07 sec)

mysql> select regr_slope(y,x) , regr_intercept(y,x) from test;
+--------------------+----------------------+
| regr_slope(y, x)   | regr_intercept(y, x) |
+--------------------+----------------------+
| 0.6853448275862069 |    5.512931034482759 |
+--------------------+----------------------+
1 row in set (0.15 sec)

[doris-robot]

Thank you for your contribution to Apache Doris.
Don't know what should be done next? See How to process your PR

Since 2024-03-18, the Document has been moved to doris-website.
See Doris Document.

[zhiqiang-hhhh]

@Yoruet Null property of regr_slope and regr_intercept should be AlwaysNullable on FE, so need to do some modification on be.

[zhiqiang-hhhh]

@Yoruet You can take this pr #40945 as a reference.

[zhiqiang-hhhh]

@Yoruet I add more comments with detail information.

[zhiqiang-hhhh]

Yoruet#1

@Yoruet I did some fix on regr_intercept on your branch, you can merge the pull request and do modification on other function. Test case should be taked.

Reference https://dbfiddle.uk/MmKrIU9r

[zhiqiang-hhhh]

Null value is not processed correctly.

[Yoruet]

Null value is not processed correctly.

Can you provide corresponding test samples to me

[zhiqiang-hhhh]

Null value is not processed correctly.

Can you provide corresponding test samples to me

@Yoruet You can try test case added in this pr Yoruet#1 on regr_slope

[Yoruet]

Null value is not processed correctly.

Can you provide corresponding test samples to me

@Yoruet You can try test case added in this pr Yoruet#1 on regr_slope

Null value is not processed correctly.

Can you provide corresponding test samples to me

@Yoruet You can try test case added in this pr Yoruet#1 on regr_slope

@zhiqiang-hhhh I compared the data in the test_regr_intercept.out file with that in https://dbfiddle.uk/MmKrIU9r, and it seems that there is nothing wrong. Can you provide more detailed information?

[zhiqiang-hhhh]

Null value is not processed correctly.

Can you provide corresponding test samples to me

@Yoruet You can try test case added in this pr Yoruet#1 on regr_slope

Null value is not processed correctly.

Can you provide corresponding test samples to me

@Yoruet You can try test case added in this pr Yoruet#1 on regr_slope

@zhiqiang-hhhh I compared the data in the test_regr_intercept.out file with that in https://dbfiddle.uk/MmKrIU9r, and it seems that there is nothing wrong. Can you provide more detailed information?

The reason test_regr_intercept is correct is that I re-write the implementation of regr_itercept in that PR (I created that pull request on your branch to give you an example on how to fix regr_itercept).

You can compare the code I submitted with your implementation of regr_itercept, and do fix on regr_slope in a same way.

[Yoruet]

Null value is not processed correctly.

Can you provide corresponding test samples to me

@Yoruet You can try test case added in this pr Yoruet#1 on regr_slope

Null value is not processed correctly.

Can you provide corresponding test samples to me

@Yoruet You can try test case added in this pr Yoruet#1 on regr_slope

@zhiqiang-hhhh I compared the data in the test_regr_intercept.out file with that in https://dbfiddle.uk/MmKrIU9r, and it seems that there is nothing wrong. Can you provide more detailed information?

The reason test_regr_intercept is correct is that I re-write the implementation of regr_itercept in that PR (I created that pull request on your branch to give you an example on how to fix regr_itercept).

You can compare the code I submitted with your implementation of regr_itercept, and do fix on regr_slope in a same way.

Thanks for your help,and I have modified regr_slope in the same way

[zhiqiang-hhhh]

Make implementation of regr_slope same with regr_intecept.

[zhiqiang-hhhh]

run buildall

[doris-robot]

TeamCity be ut coverage result:
Function Coverage: 37.24% (9626/25847)
Line Coverage: 28.67% (79699/277954)
Region Coverage: 28.09% (41214/146710)
Branch Coverage: 24.70% (20973/84896)
Coverage Report: http://coverage.selectdb-in.cc/coverage/13d467650b398ef5119bdc98bd7a1ac69e676674_13d467650b398ef5119bdc98bd7a1ac69e676674/report/index.html

[zhiqiang-hhhh]

LGTM

[github-actions]

PR approved by anyone and no changes requested.

[zhiqiang-hhhh]

We need todo a refactor on all REGR_ like agg functions, to make code simple. These agg functions have same structure.
eg. #39187

[Yoruet]

We need todo a refactor on all REGR_ like agg functions, to make code simple. These agg functions have same structure. eg. #39187

Does it include only regr_slope and regr_intercept or does it include other regr_ functions such as regr_sxx

[zhiqiang-hhhh]

We need todo a refactor on all REGR_ like agg functions, to make code simple. These agg functions have same structure. eg. #39187

Does it include only regr_slope and regr_intercept or does it include other regr_ functions such as regr_sxx

This is just a comment to remind us to do a refactor in the future, does not means we need to do refactor in this pr.

[HappenLee]

@Yoruet please add doc in https://github.com/apache/doris-website thank you very mush

[Yoruet]

@Yoruet please add doc in https://github.com/apache/doris-website thank you very mush

I have already pr the doc.plz check it