SNOW-1006168: Snowpark function array_agg converts data types to string

[FlorianWilhelm]

Please answer these questions before submitting your issue. Thanks!

1. What version of Python are you using?

   Snowsight Python notebook with snowflake-snowpark-python 1.11.1

2. What operating system and processor architecture are you using?

   Snowsight Python notebook with snowflake-snowpark-python 1.11.1

3. What are the component versions in the environment (pip freeze)?

   Snowsight Python notebook with snowflake-snowpark-python 1.11.1

4. What did you do?

I used the function array_agg and a column of big ints.

import snowflake.snowpark as snowpark
from snowflake.snowpark.functions import col, array_agg

def main(session: snowpark.Session): 
   df = session.create_dataframe([[1], [2], [3], [1]], schema=["a"])
   print("Initial dtypes:", df.dtypes)
   df = df.select(array_agg("a", False).alias("result"))
   print("After array_agg dtypes:", df.dtypes) 
   return df

5. What did you expect to see?

   I hoped to get a single row and a column with dtype array<bigint>. Instead I got array<string>.

The PY output when runing the code above is:

Initial dtypes: [('A', 'bigint')]
After array_agg dtypes: [('RESULT', 'array<string>')]

Why array<string>, this is crazy, but here is my number! as string? 🎤

[sfc-gh-yixie]

Yes we now use json string for arrays. A feature is under development for typed arrays.