Directly opening the output files through h5py

[Z10Frank]

In a few cases it can be useful to directly access the data you need using the h5py library, e.g. if the system you are working on has difficulties with Jupyter or IPython and you are trying to open the data in an extremely large output file.

This code snippet is an example to extract data from a file Fields0.h5:

import h5py
import numpy as np
import matplotlib.pyplot as plt

## open the selected hdf5 file with h5py:
filename = "Fields0.h5"
f  = h5py.File(filename,"r+")

# f will be like a typical Python dictionary, organized through "keys"
# e.g. the command f.keys() gives as result: <KeysViewHDF5 ['data']>

# In this case, the command f["data"].keys() gives as result the list of the available timesteps; 
# the length of this list can give you the number of iterations where an output was produced for this diagnostic.

# Selecting e.g. the timestep '0000001500' (if available), the command f["data"]['0000001500'].keys() 
# gives the available fields (which are those asked in the namelist for the Field diagnostic), 
# e.g. : ['Ex', 'Jx', 'Rho']

# So, if you want to extract the Ex field at timestep 1500 as numpy array, you can use for example:
myField = f["data"]['0000001500']["Ex"][()]

# Before starting to analyse the data, explore its structure by listing the available keys() 
# at each level of the HDF5 file.

[mccoys]

I usually type dict(f) or dict(f['data']) to see the content of some hdf5 group