How to access a Data asset via R function in Azure Notebook

Question

When I tried to access a data asset from R function in Azure Notebook, I failed. But I can access the same data asset in the same Azure Notebook with a python function. My intention is to initialize a mixed Python-R programming script.

Details: Conda environment Python 3.8 - AzureML. All the packaged mentioned below were installed.

## import rpy2 to use R functions
from rpy2.robjects.packages import importr

## import R package 
dt = importr('data.table')

## import Azure packages
from azure.ai.ml import MLClient
from azure.identity import DefaultAzureCredential
from azure.ai.ml.entities import Data
from azure.ai.ml.constants import AssetTypes
import azureml.fsspec


credential = DefaultAzureCredential()

ml_client = MLClient(
    credential=credential,
    subscription_id="*****************",
    resource_group_name="*************",
    workspace_name="*********",
)

## data asset info
data_asset = ml_client.data.get(name="*******", version=**)
print(f"Data asset URI: {data_asset.path}")

When I tried to use a function fread from R-data.table, one error message appeared:

df = dt.fread(data_asset.path)
df.head()

Error Message: RRuntimeError: Error: File 'azureml://subscriptions/....../datastores/....../paths/source/...csv' does not exist or is non-readable.

But when I used a function from Python in the same Azure notebook, I can read the same '.csv' file.

df = pd.read_csv(data_asset.path)
df.head()

Apparently, the data asset is accessible from this Azure notebook. Why cannot I read it through R function? I guess the problem is about the Runtime and Credential, because A "RRuntime" was created for R function in this Python Runtime. If it is the problem, does that mean I should pass the Credential to the "RRuntime"? could someone help me out?

Thank you in advance!