def __init__(self, path):

folder = []

if type(path) == h5py._hl.group.Group: #read file

folder = path.values()

else: #directly load file

folder = [os.path.join(path, f) for f in os.listdir(path) if (os.path.isfile(os.path.join(path, f)) and ('.DS_Store' not in f))]

#folder is a list that contains path of each file in the folder

#delet the hidden folder '.DS_Store' in MacOS

for filename in folder:

type_check = self.load_check(filename)

if type_check == None:

print "Data name: %s" % hspy.load(filename).metadata.General.original_filename

type_check = str(input("Please input the datatype of the data (SURVEY, HAADF, EELS, EDS, etc...): "))

print "\n"

# let users'input always is uppercase

type_check = type_check.upper()

if type_check == 'SURVEY':

if type(filename) == h5py._hl.group.Group:

self.survey = self.read_signal_from_EMDgroup(filename)

else:

self.survey = hspy.load(filename)

if type_check == 'HAADF':

if type(filename) == h5py._hl.group.Group:

self.haadf = self.read_signal_from_EMDgroup(filename)

else:

self.haadf = hspy.load(filename)

if type_check == 'EELS_L':

if type(filename) == h5py._hl.group.Group:

self.eels_l = self.read_signal_from_EMDgroup(filename)

else:

self.eels_l = hspy.load(filename)

if type_check == 'EELS_H':

if type(filename) == h5py._hl.group.Group:

self.eels_h = self.read_signal_from_EMDgroup(filename)

else:

self.eels_h = hspy.load(filename)

if type_check == 'EDS':

if type(filename) == h5py._hl.group.Group:

self.eds = self.read_signal_from_EMDgroup(filename)

else:

self.eds = hspy.load(filename)

def load_check(self, filename):

"""Function to check what type of data is contained in a file to be loaded, if possible."""

#read file from hdf5 file

data_type = ''

if type(filename) == h5py._hl.group.Group:

loaded_data = filename['data']

if loaded_data.attrs.get('type'): #check if the attribute 'type' attached on the data

data_type = loaded_data.attrs.get('type')

else:

data_type = None

#directly load file

elif type(filename) == str:

loaded_data = hspy.load(filename)

if loaded_data.metadata.Signal.signal_type == "EELS":

if loaded_data.axes_manager.signal_axes[0].offset < 500: #check if EELS is high/low-loss

data_type = 'EELS_L'

else:

data_type = 'EELS_H'

elif (loaded_data.metadata.Signal.signal_type == "EDS_SEM") or (loaded_data.metadata.Signal.signal_type == "EDS_TEM"):

data_type = 'EDS'

else:

data_type = None

return (data_type)

def read_signal_from_EMDgroup(self, group):

#extract essential data:

data = group.get('data')[...]#convert hdf5 dataset to numpy array

record_by = group['data'].attrs.get('type', '').upper()

#read data into hyperspy as Image, Spectrum or Signal:

if record_by in ['SURVEY', 'HAADF', 'IMAGE']:

signal = hspy.signals.Image(data)

elif record_by in ['EELS', 'EELS_L', 'EELS_H', 'EDS', 'SPECTRUM']:

signal = hspy.signals.Spectrum(data)

else:

signal = hspy.signals.Signal(data)

#set signal properties by iterating over all dimensions:

for i in range(len(data.shape)):

dim = group.get('dim{}'.format(i+1))

signal.axes_manager[i].name = dim.attrs.get('name', '')

string = dim.attrs.get('navigate', '')

if string == 'True':

signal.axes_manager[i].navigate = True

elif string == 'False':

signal.axes_manager[i].navigate = False

signal.axes_manager[i].offset = float(dim.attrs.get('offset', ''))

signal.axes_manager[i].scale = float(dim.attrs.get('scale', ''))

signal.axes_manager[i].size = int(dim.attrs.get('size', ''))

units = re.findall('[^_\W]+', dim.attrs.get('units', ''))

signal.axes_manager[i].units = ''.join(units)

#extract metadata:

group_metadata = {}

for key, value in group.attrs.items():

group_metadata[key] = value

signal.metadata.General.title = record_by

"""Container class for experimental data associated with a particular experiment."""

name = "experimental_data"

def __init__(self, root):

if type(root) == str: #if root is a folder path on disk

AC_folders = []

for (dir, _, _) in os.walk(root):

if not '.DS_Store' in dir: #delet the hidden folder '.DS_Store' in MacOS

if os.path.isdir(dir):

AC_folders.append(dir)

del AC_folders[0] #Without this line, the list will start with 'root' rather than 'root/dir'

#5AC_folders is a list contains physical pathnames of data on disk

elif type(root) == list: #if root is a list of group of HDF5 file

AC_folders = root

#AC_folders is a list contains group of ac1, ac2 etc. in HDF5 file

#Load class Acquisition(object)

n = 1

self.ac = {}

for f in AC_folders:

self.ac[n] = Acquisition(f)

n += 1 #'n' is the keys of the dictionary 'ac'. e.g. as follow

#ac[1] = Acquisition(folder1)

def load_experimental_data(self, exp_data):

"""Function to load experimental data to the experiment.

Example: Experiment.load_experimental_data(experimental_data)"""

self.e_data = exp_data

def save(self, filename, e_data):

"""Function to save all data associated with an experiment into group '/experimental_data'."""

f = h5py.File(filename, 'w')

f.attrs['version_major'] = 0 #EMD format required

f.attrs['version_minor'] = 2 #EMD format required

grp_e = f.create_group('experimental_data')

grp_m = f.create_group('microscope') #EMD format recommended

grp_s = f.create_group('sample') #EMD format recommended

grp_u = f.create_group('user') #EMD format recommended

grp_c = f.create_group('comments') #EMD format recommended

for ac_key, ac_val in e_data.ac.items():

ac_grp = grp_e.create_group('ac{}'.format(ac_key)) #create acqusition group named as 'ac1', 'ac2' etc.

for acs_key, acs_val in ac_val.__dict__.items():

axes_data = acs_val.axes_manager

axes_dict = acs_val.axes_manager.as_dictionary()

g = ac_grp.create_group(acs_key) #create groups named by variable names such as 'haadf, eels_h...'

g.attrs['emd_group_type'] = 1 #EMD format required

data = g.create_dataset('data', data = acs_val.data) #create main dataset 'data'

data.attrs['type'] = acs_key #.data convert data to numpy array

for k in range(len(axes_dict)):

dim = 'dim{}'.format(k+1) #name variable 'dim' as 'dim1, dim2, dim3...'

dset = g.create_dataset(dim, data = axes_data[k].axis) #create a dataset for axis

for keys, vals in axes_data[k].get_axis_dictionary().items():

if keys == 'units': #??? units still not quite fit the EMD specification

dset.attrs[keys] = '[{}]'.format('_'.join(list(str(vals))))

else:

dset.attrs[keys] = str(vals)

def read(self, filename):

self.r = h5py.File(filename, 'r')

def func_1(path): # define a callable func for .visit(func)

if 'experimental_data' in path:

return path

p = self.r.visit(func_1) #find the path of the group 'experimental_data'

AC_folders = self.r[p].values() #it's a list of groups within the group 'experimental_data'