README.md

hdfio

A set of high-level utilities for working with HDF5.

This package is not meant to expose anywhere near the full capabilities of HDF5. For that, see the rhdf5 and hdf5r packages (we actually use hdf5r internally). The goal of this package is to try to make HDF5 as simple to use as read.csv() and write.csv(), but with the added benefits of using binary file formats.

Our main focus is on storing and reading dataframes. We call these "h5df" as in "h5 dataframe". I understand that this is annoying and difficult to convince your fingers to type if you are familiar with HDF5, but it's too good of a name to pass up. Right now we support reading from two kinds of formats written by python's pandas (really pytables), with some restrictions (no strings when format=fixed). We have full support for a format that is good for working with R, and should soon have a format that is useful if the goal is to regularly share data between python and R.

The current documentation is a train wreck, but we're working on it.

Installation



RBigData/hdfio documentation built on May 21, 2019, 1:20 p.m.