Chunked HDF vs Memmapped Array I/O Comparison: http://stackoverflow.com/questions/27710245/is-there-an-analysis-speed-or-memory-usage-advantage-to-using-hdf5-for-large-arr/27713489#27713489

chunked_hdf.py

import sys
import h5py

def main():
    data = read()

    if sys.argv[1] == 'x':
        x_slice(data)
    elif sys.argv[1] == 'z':
        z_slice(data)

def read():
    f = h5py.File('/tmp/test.hdf5', 'r')
    return f['seismic_volume']

def z_slice(data):
    return data[:,:,0]

def x_slice(data):
    return data[0,:,:]

main()

memmapped_array.py

import numpy as np
import sys

def main():
    data = read()

    if sys.argv[1] == 'x':
        x_slice(data)
    elif sys.argv[1] == 'z':
        z_slice(data)

def read():
    big_binary_filename = '/data/nankai/data/Volumes/kumdep01_flipY.3dv.vol'
    shape = 621, 4991, 2600
    header_len = 3072

    data = np.memmap(filename=big_binary_filename, mode='r', offset=header_len,
                     order='F', shape=shape, dtype=np.uint8)
    return data

def z_slice(data):
    dat = np.empty(data.shape[:2], dtype=data.dtype)
    dat[:] = data[:,:,0]
    return dat

def x_slice(data):
    dat = np.empty(data.shape[1:], dtype=data.dtype)
    dat[:] = data[0,:,:]
    return dat

main()

write_chunked_hdf_from_memmapped_array.py

import geoprobe
import h5py

# I'm using a library here, but "vol.data" is just a memmapped array.
vol = geoprobe.volume('/data/nankai/data/Volumes/kumdep01_flipY.3dv.vol')

with h5py.File('/tmp/test.hdf5', 'w') as f:
    dset = f.create_dataset('seismic_volume', shape=vol.data.shape,
                            dtype=vol.data.dtype, chunks=True)

    # Can't fit the entire file in memory...
    for i in range(vol.nz):
        dset[:,:,i] = vol.data[:,:,i]