You are viewing a plain text version of this content. The canonical link for it is here.
Posted to reviews@iotdb.apache.org by GitBox <gi...@apache.org> on 2021/02/17 04:05:08 UTC

[GitHub] [iotdb] qiaojialin opened a new issue #2687: BugAbout NaN

qiaojialin opened a new issue #2687:
URL: https://github.com/apache/iotdb/issues/2687


   version 0.11.2:
   
   ```
   INSERT INTO root.happy.device1.sensor1(timestamp,temperature) values(7925, NaN)
   ```
   
   Msg: 500: For input string: "NaN"


----------------------------------------------------------------
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.

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



[GitHub] [iotdb] weizihan0110 commented on issue #2687: Bug about inserting NaN

Posted by GitBox <gi...@apache.org>.
weizihan0110 commented on issue #2687:
URL: https://github.com/apache/iotdb/issues/2687#issuecomment-781932021


   Because it's converted into the Long type at org.apache.iotdb.tsfile.encoding.encoder.FloatEncoder#convertDoubleToLong.
   And my doubts are why there's no something like DoubleRleEncoder?
   ![image](https://user-images.githubusercontent.com/15262869/108481281-f7bd2980-72d2-11eb-809b-aad48d00dcef.png)
   


----------------------------------------------------------------
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.

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



[GitHub] [iotdb] weizihan0110 commented on issue #2687: Bug about inserting NaN

Posted by GitBox <gi...@apache.org>.
weizihan0110 commented on issue #2687:
URL: https://github.com/apache/iotdb/issues/2687#issuecomment-781933282


   Another question: should we support NaN or NULL for BOOLEAN, INT32, INT64?


----------------------------------------------------------------
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.

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



[GitHub] [iotdb] weizihan0110 commented on issue #2687: Bug about inserting NaN

Posted by GitBox <gi...@apache.org>.
weizihan0110 commented on issue #2687:
URL: https://github.com/apache/iotdb/issues/2687#issuecomment-780651552


   To solve first problem in version 0.11.2, just Cherry Pick a70701220bdad25be18bb58ee8253c536aed4367 from master.


----------------------------------------------------------------
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.

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



[GitHub] [iotdb] HTHou commented on issue #2687: Bug about inserting NaN

Posted by GitBox <gi...@apache.org>.
HTHou commented on issue #2687:
URL: https://github.com/apache/iotdb/issues/2687#issuecomment-781868324


   Another problem...
   ```
   IoTDB> CREATE TIMESERIES root.happy.device1.sensor1.temperature WITH DATATYPE=DOUBLE, ENCODING=RLE
   IoTDB> INSERT INTO root.happy.device1.sensor1(timestamp,temperature) values(7925, NaN)
   IoTDB> select * from root
   +-----------------------------+--------------------------------------+
   |                         Time|root.happy.device1.sensor1.temperature|
   +-----------------------------+--------------------------------------+
   |1970-01-01T08:00:07.925+08:00|                                   NaN|
   +-----------------------------+--------------------------------------+
   IoTDB> flush
   IoTDB> select * from root
   +-----------------------------+--------------------------------------+
   |                         Time|root.happy.device1.sensor1.temperature|
   +-----------------------------+--------------------------------------+
   |1970-01-01T08:00:07.925+08:00|                                   0.0|
   +-----------------------------+--------------------------------------+
   ```


----------------------------------------------------------------
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.

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



[GitHub] [iotdb] qiaojialin commented on issue #2687: Bug about inserting NaN

Posted by GitBox <gi...@apache.org>.
qiaojialin commented on issue #2687:
URL: https://github.com/apache/iotdb/issues/2687#issuecomment-780349303


   Todo:
   (1) fix the 500 exception in 0.11.2
   (2) fix the print 0.0 in 0.12.0-SNAPSHOT


----------------------------------------------------------------
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.

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



[GitHub] [iotdb] HTHou edited a comment on issue #2687: Bug about inserting NaN

Posted by GitBox <gi...@apache.org>.
HTHou edited a comment on issue #2687:
URL: https://github.com/apache/iotdb/issues/2687#issuecomment-781128459






----------------------------------------------------------------
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.

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



[GitHub] [iotdb] HTHou commented on issue #2687: Bug about inserting NaN

Posted by GitBox <gi...@apache.org>.
HTHou commented on issue #2687:
URL: https://github.com/apache/iotdb/issues/2687#issuecomment-781972673


   > And my doubts are why there's no something like DoubleRleEncoder?
   
   It seems that the way of encoding DOUBLE and FLOAT data with RLE and TS_2DIFF is converting to long and int first. And then encode the long and int values.
   
   
   


----------------------------------------------------------------
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.

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



[GitHub] [iotdb] HTHou commented on issue #2687: Bug about inserting NaN

Posted by GitBox <gi...@apache.org>.
HTHou commented on issue #2687:
URL: https://github.com/apache/iotdb/issues/2687#issuecomment-781955835


   > Another question: should we support NaN or NULL for BOOLEAN, INT32, INT64?
   
   There is a jira issue about this. https://issues.apache.org/jira/projects/IOTDB/issues/IOTDB-1158?filter=allissues
   
   In my opinion, I agree to support `NULL` for all value types. 
   As for `NaN`, I think it can only be float or double type, since this is only defined in JDK classes of Double and Float.


----------------------------------------------------------------
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.

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



[GitHub] [iotdb] weizihan0110 commented on issue #2687: Bug about inserting NaN

Posted by GitBox <gi...@apache.org>.
weizihan0110 commented on issue #2687:
URL: https://github.com/apache/iotdb/issues/2687#issuecomment-786430095


   > > And my doubts are why there's no something like DoubleRleEncoder?
   > 
   > It seems that the way of encoding DOUBLE and FLOAT data with RLE and TS_2DIFF is converting to long and int first. And then encode the long and int values.
   
   But I still don't understand why something like DoubleRleEncoder is not introduced... 


----------------------------------------------------------------
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.

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



[GitHub] [iotdb] HTHou commented on issue #2687: Bug about inserting NaN

Posted by GitBox <gi...@apache.org>.
HTHou commented on issue #2687:
URL: https://github.com/apache/iotdb/issues/2687#issuecomment-781128459


   I find the reason may be related with the encoding type.
   ```
   IoTDB> INSERT INTO root.happy.device1.sensor1(timestamp,temperature) values(7925, NaN)
   Msg: The statement is executed successfully.
   IoTDB> select * from root
   +-----------------------------+--------------------------------------+
   |                         Time|root.happy.device1.sensor1.temperature|
   +-----------------------------+--------------------------------------+
   |1970-01-01T08:00:07.925+08:00|                                   NaN|
   +-----------------------------+--------------------------------------+
   Total line number = 1
   ```


----------------------------------------------------------------
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.

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