Functions: I can read parquet file from my C drive. However, I can't write it back. What I am doing wrong?

[mraguth]

I am a fan of Autogen.
The fundamental question here is... can I write Autogen output back to my local drive? Either via in built functions or my own functions.

I tried with my own function. I can read but I can't write.... Any help would be greatly appreciated.

This is my code:

My functions:

def prq_read(filename: str) -> pd.DataFrame:
    """
    Reads a Parquet file from the specified directory and returns a pandas DataFrame.

    Args:
        filename (str): The name of the Parquet file.

    Returns:
        pd.DataFrame: The data read from the Parquet file.
    """
    parquet_file = f"C:/parquet/{filename}"
    dataframe = pd.read_parquet(parquet_file)
    return dataframe
    #return f"{dataframe} {message}"


def prq_save(dataframe):
    """
    Writes a pandas DataFrame to a Parquet file.
    Args:
        dataframe (pd.DataFrame): The DataFrame to be written.
    Returns:
        str: The name of the output Parquet file.
    """
    table = pa.Table.from_pandas(dataframe)
    pq.write_table(table, 'c:/parquet/result.parquet')
    return dataframe

llm_config:

llm_config = {
    "config_list": config_list,
    "seed": 42,
    "functions":[
        {
            "name": "prq_read",
            "description": "Reads a Parquet file from the specified directory and returns a pandas DataFrame.",
            "parameters": {
                "type": "object",
                "properties": {
                    "filename":{
                        "type": "string",
                        "description": "Reads a Parquet file from the specified directory and returns a pandas DataFrame."
                    }
                },
                "required":["filename"]
            }
        },
        {
            "name": "prq_save",
            "description": "Read pandas dataframe and convert it to parquet file and save the parquet file.",
            "parameters": {
                "type": "object",
                "properties": {
                    "dataframe":{
                        "type": "string",
                        "description": "Read pandas dataframe and convert it to parquet file and save the parquet file in the specified directory"
                    }
                },
                "required":["dataframe"]
            }
        }
    ]
}

coder = autogen.AssistantAgent(
    name="chatbot",
    system_message="For coding tasks, only use the functions you have been provided with. Reply TERMINATE when the task is done.",
    llm_config=llm_config,
)

coder.register_function(
    function_map={
        "prq_read": prq_read,
        "prq_save": prq_save,
    }
)

My call:

message = """
1) Read murali.parquet file and convert it to pandas dataframe.
2) Display top 3 rows.
3) convert the pandas dataframe to parquet format and save as murali2.parquet in 'c:/parquet'.
4) Execute the above process and TERMINANTE.
"""

with Cache.disk() as cache:
    await user_proxy.a_initiate_chat(  # noqa: F704
    manager,
    message=message)

Error I am getting at prq_save function (or this function helps in writing the data back to my C drive)

--------------------------------------------------------------------------------

>>>>>>>> EXECUTING FUNCTION prq_save...
chatbot (to chat_manager):

***** Response from calling function "prq_save" *****
Error: 'str' object has no attribute 'columns'
*****************************************************