[Python] Unable to use dask-sql due to 'dask_expr.io' module

Aim:

1. Read data from Parquet files
2. Register each df as Table
3. Use dask-sql to join & query from the table

Here are the installation step:

• pip install --force-reinstall --no-cache-dir "dask[complete]" dask-sql
• pip install dask-expr

Installed versions:

import dask
import dask_sql

print(dask.__version__)
print(dask_sql.__version__)
# output:
> 2025.5.1
> 2024.5.0


Steps to reproduce error with sample CSV file:

import dask.dataframe as dd
from dask_sql import Context

tmp_df = dd.read_csv("./tmp.csv")

c = Context()
c.create_table("tmp", df)


---------------------------------------------------------------------------
ModuleNotFoundError Traceback (most recent call last)
Cell In[2], line 2
1 c = Context()
----> 2 c.create_table("tmp", tmp_df)

File ~\AppData\Roaming\Python\Python312\site-packages\dask_sql\context.py:266, in Context.create_table(self, table_name, input_table, format, persist, schema_name, statistics, gpu, **kwargs)
264 try:
265 if dd._dask_expr_enabled():
--> 266 from dask_expr.io.parquet import ReadParquet
268 dask_filepath = None
269 operations = input_table.find_operations(ReadParquet)

ModuleNotFoundError: No module named 'dask_expr.io'


Note: even after installing pip install dask-expr I am getting the same error.

Continue reading...