Table.export('path/myfile.tsv.bgz')


#1

Hi,

I have an issue using the table.export() function as it does not save any file. I tried with different paths and folders with different permits and I consistently get the same behavior. There are no errors displayed and I even get the merge time metric.

On the other hand, the following works: myTable=table.to_pandas() and then myTable.to_csv(). However, I think that using export may allow saving larger files.

Any suggestion of that may be happening here?


#2

I got it. It saves the files into the hdfs dfs Hadoop file system.


#3

On a HDFS-enabled cluster, the default file scheme is hdfs. You can write to file:///path/in/unix/file/system though


#4

or gs://, or s3://, …


#5

Thank you, file:// and s3:// work great!