When I use "pd.read_parquet(path,engine='pyarrow??)?? to load my dataset,the 
loaded dataset appears where one piece of data is copied into two??but 
fastparquet will not.

Reply via email to