'ValueError' Upon SQL Cell Execution

Problems and bugs
1

Running Datalore On-Prem, Docker-Compose

Upon execution of a SQL Cell, ValueError is thrown.

Begin Traceback Block:

Traceback (most recent call last):
  at block 2, line 1
  at block 3, line 7, in _datalore_internal_python_execute_sql_query(sql_query, database_id, user_id, parameters)
  at /opt/anaconda3/envs/minimal/lib/python3.8/site-packages/pandas/core/frame.pyline 694, in __init__(self, data, index, columns, dtype, copy)
  at /opt/anaconda3/envs/minimal/lib/python3.8/site-packages/pandas/core/internals/construction.pyline 483, in nested_data_to_arrays(data, columns, index, dtype)
  at /opt/anaconda3/envs/minimal/lib/python3.8/site-packages/pandas/core/internals/construction.pyline 807, in to_arrays(data, columns, dtype)
  at /opt/anaconda3/envs/minimal/lib/python3.8/site-packages/pandas/core/internals/construction.pyline 909, in _finalize_columns_and_data(content, columns, dtype)
ValueError: 13 columns passed, passed data had 12 columns

I have ruled out the query, as it executes fine on DBBeaver & DBForge. I have attempted simplistic queries that don’t involve as many join statements. However, same error is thrown.

Has anyone else had this issue?

2

It could be a bug on our side. Does it happen when working with a MySQL database? Please send me DESCRIBE output for this table and, if possible, dump or some sample data (via DM) to reproduce the issue.

Thank you!

Igor

3

Hi igro,

I encounter the same issue.
I can simply take the following steps to reproduce it:

  • create a SQL cell
  • select the connection with a clickhouse db
  • run SQL “select count(1)”
    and then the error popup:
Traceback (most recent call last):
  at block 4, line 1
  at block 4, line 9, in _datalore_internal_python_execute_sql_query(sql_query, database_id, user_id, parameters)
  at /opt/python/envs/minimal/lib/python3.8/site-packages/pandas/core/frame.py, line 694, in __init__(self, data, index, columns, dtype, copy)
  at /opt/python/envs/minimal/lib/python3.8/site-packages/pandas/core/internals/construction.py, line 483, in nested_data_to_arrays(data, columns, index, dtype)
  at /opt/python/envs/minimal/lib/python3.8/site-packages/pandas/core/internals/construction.py, line 807, in to_arrays(data, columns, dtype)
  at /opt/python/envs/minimal/lib/python3.8/site-packages/pandas/core/internals/construction.py, line 909, in _finalize_columns_and_data(content, columns, dtype)
ValueError: 1 columns passed, passed data had 0 columns

below is the screenshot:

image
image2156×560 82 KB

4

@leonli, thank you for the report!

I managed to reproduce this error locally – it seems there is an issue with the Clickhouse driver. Forwarded this report to the dev team, will keep you posted!

Best regards,
Igor

1 Like
5

We found an issue with the returned types (String comes as a Numeric sql type). We added a proper support for this behavior at our side - it is already implemented at our cloud instance (datalore.jetbrains.com) and the fix will be also included into the next planned Enterprise release (2023.2, mid April).

Thanks again for the report!

6

I have the same problem with an Enterprise deployment of 2023.2 using a Trino connection with a MariaDB connector whenever the SELECT contains a TINYINT column. Other connectors and other data types work fine though, even casting the column from TINYINT to SMALLINT is fine.
Also ruled out the query, it works fine when executed through JDBC.

2 Likes
7

@domnikl, thank you for reporting this issue! I managed to reproduce the error and forwarded the report to the dev team.

1 Like