h2o.importFolder
should ignore _SUCCESS
and .crc
files by default
#16409
Labels
h2o.importFolder
should ignore _SUCCESS
and .crc
files by default
#16409
Is your feature request related to a problem? Please describe.
When you export data (e.g., parquets) from Spark you will generally have a
_SUCCESS
file and.crc
(checksum) files in the directory.h2o
does not like these files.h2o
should, like Spark does, ignore them when importing a directory. Otherwise this error is thrown:Describe the solution you'd like
h2o.importFolder(...)
runs without error even if the directory contains a_SUCCESS
file or.crc
files. Note that evenh2o.exportFile
by default writes.crc
ifformat = "parquet"
.Describe alternatives you've considered
One option is to use the
pattern
argument, but that's somewhat of a nuisance.The text was updated successfully, but these errors were encountered: