Importing a File

Unlike the upload function, which is a push from the client to the server, the import function is a parallelized reader and pulls information from the server from a location specified by the client. The path is a server-side path. This is a fast, scalable, highly optimized way to read data. H2O pulls the data from a data store and initiates the data transfer as a read operation.

Refer to the Supported File Formats topic to ensure that you are using a supported file type.

Note: When parsing a data file containing timestamps that do not include a timezone, the timestamps will be interpreted as UTC (GMT). You can override the parsing timezone using the following:

  • R: h2o.setTimezone("America/Los Angeles")
  • Python: h2o.cluster().timezone = "America/Los Angeles"
# To import airlines file from H2O’s package:
> library(h2o)
> h2o.init()
> irisPath <- "https://s3.amazonaws.com/h2o-airlines-unpacked/allyears2k.csv"
> iris.hex <- h2o.importFile(path = irisPath, destination_frame = "iris.hex")

# To import from S3:
> library(h2o)
> h2o.init()
> airlinesURL <- "https://s3.amazonaws.com/h2o-airlines-unpacked/allyears2k.csv"
> airlines.hex <- h2o.importFile(path = airlinesURL, destination_frame = "airlines.hex")

# To import from HDFS, you must include the node name:
> library(h2o)
> h2o.init()
> airlinesURL <- "hdfs://node-1:/user/smalldata/airlines/allyears2k_headers.zip"
> airlines.hex <- h2o.importFile(path = airlinesURL, destination_frame = "airlines.hex")
# Import a file from S3:
>>> import h2o
>>> h2o.init()
>>> airlines = "http://s3.amazonaws.com/h2o-public-test-data/smalldata/airlines/allyears2k_headers.zip"
>>> airlines_df = h2o.import_file(path=airlines)

# Import a file from HDFS, you must include the node name:
>>> import h2o
>>> h2o.init()
>>> airlines = "hdfs://node-1:/user/smalldata/airlines/allyears2k_headers.zip"
>>> airlines_df = h2o.import_file(path=airlines)