Re: [PR] Implement PARQUET-2249: Introduce IEEE 754 total order [arrow-rs]

["HuaHuaY (via GitHub)" <gi...@apache.org>]

HuaHuaY commented on code in PR #9619:
URL: https://github.com/apache/arrow-rs/pull/9619#discussion_r3666586303


##########
parquet/src/file/statistics.rs:
##########
@@ -139,6 +139,8 @@ pub(crate) fn from_thrift_page_stats(
                 .transpose()?;
             // Generic distinct count (count of distinct values occurring)
             let distinct_count = stats.distinct_count.map(|value| value as u64);
+            // Generic nan count for floating point types
+            let nan_count = stats.nan_count.map(|value| value as u64);

Review Comment:
   Should it throw an error when the count is negative, just like with `null_count`? Would it be better to treat negative values ​​in both cases as "missing"? I noticed that Arrow C++ doesn't handle negative values, so I try to follow the Rust implementation behavior.



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: github-unsubscribe@arrow.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org