Hello, 

I am Yunxiang Su from School of Software, Tsinghua University, supervised by 
Shaoxu Song. Recently, our industrial partners report some bugs in Percentile 
and Quantile. I fixed these bugs, and then add a new UDF “Cluster" for time 
series subsequence clustering. To be specific, 
- Fix out-of-bounds errors in “Percentile” UDF; 
- Fix type conversion errors in “Quantile” UDF; 
- Add a new useful UDF, named “Cluster”, which takes a **single input time 
series**, splits it into **non-overlapping** contiguous subsequences (windows) 
of fixed length `l`, and clusters those subsequences into `k` groups.

I have proposed the PRs to master (https://github.com/apache/iotdb/pull/17375), 
dev/1.3 branch (https://github.com/apache/iotdb/pull/17447) of IoTDB, and the 
corresponding docs (https://github.com/apache/iotdb-docs/pull/1072). 

Please check my PRs, and contact me if there are any issues. Thanks.

Sincerely, 
Yunxiang Su

Reply via email to