[Bug] doris Function round imprecision

[longzmkm]

Search before asking

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

Version

doris-2.1.0-rc11-91efb6a43d

What's Wrong?

CREATE TABLE test_double (
report_date date,
v double NULL
) engine=OLAP
DUPLICATE KEY(report_date)
DISTRIBUTED BY HASH(report_date) BUCKETS 1
PROPERTIES (
"replication_allocation" = "tag.location.default: 1",
"in_memory" = "false",
"storage_format" = "V2",
"disable_auto_compaction" = "false"
);

INSERT INTO test_double (report_date, v) VALUES
('2024-03-19', NULL),
('2024-03-19', '2'),
('2024-03-19', '0'),
('2024-03-19', '0'),
('2024-03-19', '0'),
('2024-03-19', '0'),
('2024-03-19', '1'),
('2024-03-19', NULL),
('2024-03-19', '0.5'),
('2024-03-19', NULL),
('2024-03-19', NULL),
('2024-03-19', NULL),
('2024-03-19', NULL),
('2024-03-19', '0'),
('2024-03-19', '1'),
('2024-03-19', '0'),
('2024-03-19', NULL);

SELECT
report_date,
ROUND(4.5,0),
round(v_total,0),
v_total
from
(
SELECT
test_double.report_date,
sum(COALESCE(v,0)) as "v_total"
from test_double
GROUP by
test_double.report_date
)a;

why round(v_total,0) = 4 but 5 is true

What You Expected?

why round(v_total,0) = 4 but 5 is true

How to Reproduce?

No response

Anything Else?

No response

Are you willing to submit PR?

• Yes I am willing to submit a PR!

Code of Conduct

• I agree to follow this project's Code of Conduct

[Mryange]

#31583
Previously, our rounding logic was to round to the nearest even number,now,change

[longzmkm]

#31583 Previously, our rounding logic was to round to the nearest even number,now,change

Which version will it be released in， about 2.1.X

[github-actions]

We're closing this issue because it hasn't been updated in a while.
If you'd like to revive this issue, please reopen it and feel free a maintainer to remove the Stale tag!