def read_surface(fname, verbose=None):
"""Load a Freesurfer surface mesh in triangular format
Parameters
----------
fname : str
The name of the file containing the surface.
verbose : bool, str, int, or None
If not None, override default verbose level (see mne.verbose).
Returns
-------
rr : array, shape=(n_vertices, 3)
Coordinate points.
tris : int array, shape=(n_faces, 3)
Triangulation (each line contains indexes for three points which
together form a face).
"""
with open(fname, "rb", buffering=0) as fobj: # buffering=0 for np bug
magic = _fread3(fobj)
if (magic == 16777215) or (magic == 16777213): # Quad file or new quad
nvert = _fread3(fobj)
nquad = _fread3(fobj)
coords = np.fromfile(fobj, ">i2", nvert * 3).astype(np.float)
coords = coords.reshape(-1, 3) / 100.0
quads = _fread3_many(fobj, nquad * 4)
quads = quads.reshape(nquad, 4)
#
# Face splitting follows
#
faces = np.zeros((2 * nquad, 3), dtype=np.int)
nface = 0
for quad in quads:
if (quad[0] % 2) == 0:
faces[nface] = quad[0], quad[1], quad[3]
nface += 1
faces[nface] = quad[2], quad[3], quad[1]
nface += 1
else:
faces[nface] = quad[0], quad[1], quad[2]
nface += 1
faces[nface] = quad[0], quad[2], quad[3]
nface += 1
elif magic == 16777214: # Triangle file
create_stamp = fobj.readline()
_ = fobj.readline() # analysis:ignore
vnum = np.fromfile(fobj, ">i4", 1)[0]
fnum = np.fromfile(fobj, ">i4", 1)[0]
#raise RuntimeError
coords = np.fromfile(fobj, ">f4", vnum * 3).reshape(vnum, 3)
faces = np.fromfile(fobj, ">i4", fnum * 3).reshape(fnum, 3)
else:
raise ValueError("%s does not appear to be a Freesurfer surface"
% fname)
logger.info('Triangle file: %s nvert = %s ntri = %s'
% (create_stamp.strip(), len(coords), len(faces)))
coords = coords.astype(np.float) # XXX: due to mayavi bug on mac 32bits
return coords, faces
def _read_frame(self, file_num, frame_num, raw_flag):
pdict = self.flist[file_num]
with open(pdict['fname'], 'rb') as fptr:
num_data = np.fromfile(fptr, dtype='i4', count=1)[0]
accum = [pdict['ones_accum'], pdict['multi_accum']]
offset = [0, 0]
size = [0, 0]
if frame_num == 0:
size = [accum[0][frame_num], accum[1][frame_num]]
else:
offset = [accum[0][frame_num-1], accum[1][frame_num-1]]
size[0] = accum[0][frame_num] - accum[0][frame_num - 1]
size[1] = accum[1][frame_num] - accum[1][frame_num - 1]
fptr.seek(1024 + num_data*8 + offset[0]*4, 0)
place_ones = np.fromfile(fptr, dtype='i4', count=size[0])
fptr.seek(1024 + num_data*8 + accum[0][-1]*4 + offset[1]*4, 0)
place_multi = np.fromfile(fptr, dtype='i4', count=size[1])
fptr.seek(1024 + num_data*8 + accum[0][-1]*4 + accum[1][-1]*4 + offset[1]*4, 0)
count_multi = np.fromfile(fptr, dtype='i4', count=size[1])
frame = np.zeros(pdict['num_pix'], dtype='i4')
np.add.at(frame, place_ones, 1)
np.add.at(frame, place_multi, count_multi)
frame *= pdict['geom'].unassembled_mask
if not raw_flag:
frame = self._assemble_frame(frame, file_num)
return frame
def mnist(ntrain=60000,ntest=10000,onehot=True): data_dir = os.path.join(datasets_dir,'mnist/') fd = open(os.path.join(data_dir,'train-images.idx3-ubyte')) loaded = np.fromfile(file=fd,dtype=np.uint8) trX = loaded[16:].reshape((60000,28*28)).astype(float) fd = open(os.path.join(data_dir,'train-labels.idx1-ubyte')) loaded = np.fromfile(file=fd,dtype=np.uint8) trY = loaded[8:].reshape((60000)) fd = open(os.path.join(data_dir,'t10k-images.idx3-ubyte')) loaded = np.fromfile(file=fd,dtype=np.uint8) teX = loaded[16:].reshape((10000,28*28)).astype(float) fd = open(os.path.join(data_dir,'t10k-labels.idx1-ubyte')) loaded = np.fromfile(file=fd,dtype=np.uint8) teY = loaded[8:].reshape((10000)) trX = trX/255. teX = teX/255. trX = trX[:ntrain] trY = trY[:ntrain] teX = teX[:ntest] teY = teY[:ntest] if onehot: trY = one_hot(trY, 10) teY = one_hot(teY, 10) else: trY = np.asarray(trY) teY = np.asarray(teY) return trX,teX,trY,teY
def loadlocal_mnist(images_path, labels_path):
""" Read MNIST from ubyte files.
Parameters
----------
images_path : str
path to the test or train MNIST ubyte file
labels_path : str
path to the test or train MNIST class labels file
Returns
--------
images : [n_samples, n_pixels] numpy.array
Pixel values of the images.
labels : [n_samples] numpy array
Target class labels
"""
with open(labels_path, 'rb') as lbpath:
magic, n = struct.unpack('>II',
lbpath.read(8))
labels = np.fromfile(lbpath,
dtype=np.uint8)
with open(images_path, 'rb') as imgpath:
magic, num, rows, cols = struct.unpack(">IIII",
imgpath.read(16))
images = np.fromfile(imgpath,
dtype=np.uint8).reshape(len(labels), 784)
return images, labels
def getCoincidences(fTimes, fChans, gate, radius, heraldChan):
bufRes = 156.25e-12
gate = int(gate/bufRes)
radius = int(radius/bufRes)
coin = np.zeros([8,8], dtype = np.uint64)
times = np.fromfile(fTimes, dtype = np.uint64)
chans = np.fromfile(fChans, dtype = np.uint8)
#print "len(times), len(chans)", len(times), len(chans)
for chan in range(8,16):
colIdx = np.where(chans==chan)[0]
for idx in colIdx:
#print "chans[idx]: %d"%chans[idx]
#print "few chans: ",chans[idx-3:idx+3]
#print "few times: ",times[idx-3:idx+3]
j = idx + 1
while (j < len(times)) and (chans[j]==heraldChan) and (times[j] - gate <= times[idx]):
i = idx - 1
while (i >= 0):
if (times[i] + radius >= times[idx]) and (chans[idx] != chans[i]) and chans[i] < 8:
row = chans[i]
col = chans[idx] % 8
coin[row, col] += 1
break
elif (times[i] + radius <= times[idx]):
row = heraldChan % 8 #works even if for some reason we had the rows plugged into channels 8-15 of the tagger
col = chans[idx] % 8
coin[row, col] += 1
break
i -= 1
j += 1
return coin
"""
def read_morph_data(filepath):
"""Read a Freesurfer morphometry data file.
This function reads in what Freesurfer internally calls "curv" file types,
(e.g. ?h. curv, ?h.thickness), but as that has the potential to cause
confusion where "curv" also refers to the surface curvature values,
we refer to these files as "morphometry" files with PySurfer.
Parameters
----------
filepath : str
Path to morphometry file
Returns
-------
curv : numpy array
Vector representation of surface morpometry values
"""
with open(filepath, "rb") as fobj:
magic = _fread3(fobj)
if magic == 16777215:
vnum = np.fromfile(fobj, ">i4", 3)[0]
curv = np.fromfile(fobj, ">f4", vnum)
else:
vnum = magic
_fread3(fobj)
curv = np.fromfile(fobj, ">i2", vnum) / 100
return curv
def Read(self):
#return numpy.ones((256, 819)).astype('float32'), numpy.ones(256).astype('int32')
with open(self.featureFile,"rb") as f:
dt = numpy.dtype([('numSamples',(numpy.int32,1)),('sampPeriod',(numpy.int32,1)),('sampSize',(numpy.int16,1)),('sampKind',(numpy.int16,1))])
header = numpy.fromfile(f,dt.newbyteorder('>' if self.byteOrder==ByteOrder.BigEndian else '<'),count=1)
numSamples = header[0]['numSamples']
sampPeriod = header[0]['sampPeriod']
sampSize = header[0]['sampSize']
sampKind = header[0]['sampKind']
# print 'Num samples = {}'.format(numSamples)
# print 'Sample period = {}'.format(sampPeriod)
# print 'Sample size = {}'.format(sampSize)
# print 'Sample kind = {}'.format(sampKind)
dt = numpy.dtype([('sample',(numpy.float32,sampSize/4))])
samples = numpy.fromfile(f,dt.newbyteorder('>' if self.byteOrder==ByteOrder.BigEndian else '<'),count=numSamples)
self._markDone()
if self.labelFile is None:
labels = None
else:
labels = ReadLabel(self.labelFile)
return samples[:]['sample'], labels
def _read_volume_info(fobj):
"""An implementation of nibabel.freesurfer.io._read_volume_info, since old
versions of nibabel (<=2.1.0) don't have it.
"""
volume_info = dict()
head = np.fromfile(fobj, '>i4', 1)
if not np.array_equal(head, [20]): # Read two bytes more
head = np.concatenate([head, np.fromfile(fobj, '>i4', 2)])
if not np.array_equal(head, [2, 0, 20]):
warnings.warn("Unknown extension code.")
return volume_info
volume_info['head'] = head
for key in ['valid', 'filename', 'volume', 'voxelsize', 'xras', 'yras',
'zras', 'cras']:
pair = fobj.readline().decode('utf-8').split('=')
if pair[0].strip() != key or len(pair) != 2:
raise IOError('Error parsing volume info.')
if key in ('valid', 'filename'):
volume_info[key] = pair[1].strip()
elif key == 'volume':
volume_info[key] = np.array(pair[1].split()).astype(int)
else:
volume_info[key] = np.array(pair[1].split()).astype(float)
# Ignore the rest
return volume_info
def _load_ahf_particle_block(self, f):
"""Load the particles for the next halo described in particle file f"""
ng = len(self.base.gas)
nds = len(self.base.dark) + len(self.base.star)
nparts = int(f.readline().split()[0])
if self.isnew:
if isinstance(f, file):
data = (np.fromfile(
f, dtype=int, sep=" ", count=nparts*2).reshape(nparts, 2))[:, 0]
else:
# unfortunately with gzipped files there does not
# seem to be an efficient way to load nparts lines
data = np.zeros(nparts, dtype=int)
for i in xrange(nparts):
data[i] = int(f.readline().split()[0])
if self._use_iord :
data = self._iord_to_fpos[data]
else :
hi_mask = data >= nds
data[np.where(hi_mask)] -= nds
data[np.where(~hi_mask)] += ng
else:
if isinstance(f, file):
data = np.fromfile(f, dtype=int, sep=" ", count=nparts)
else:
# see comment above on gzipped files
data = np.zeros(nparts, dtype=int)
for i in xrange(nparts):
data[i] = int(f.readline())
data.sort()
return data
def train_generator():
random.seed(seedvalue)
valtruth = list()
valdata = list()
endflag = False
for i in range(0,len(events)):
for j in range(0,len(truthfiles)):
filename = events[i]+truthfiles[j]
try:
truth = np.fromfile(filename,dtype=np.short)
truth = truth.reshape(len(truth)//wirelength,wirelength)
for k in range(0,len(datafiles[j])):
filename = events[i]+datafiles[j][k]
try:
data = np.fromfile(filename,dtype=np.short)
data = data.reshape(len(data)//wirelength,wirelength)
data = np.concatenate((padarray,data,padarray),axis=0)
for l in range(batchsize,len(truth)+1,batchsize):
if i>=len(events)-1 and j>=len(truthfiles)-1 and k>=len(datafiles[j])-1 and l>=len(truth): endflag=True
if random.uniform(0,1) > droptrainingwires or endflag:
start = l - batchsize
end = l
t = list()
for m in range(start,end): t.append((np_utils.to_categorical(truth[m],numlabels)))
d = parse_attributes(data[start:end+numattributes-1])
if random.uniform(0,1) < reservevalidation and not endflag:
valtruth.extend(t)
valdata.extend(d)
else:
yield np.asarray(d), np.asarray(t), np.asarray(valdata), np.asarray(valtruth), endflag
except IOError:
print 'No file named',filename
except IOError:
print 'No truth file for',events[i]
def read_lgal_input_fulltrees_withids(folder,lastsnap,file,verbose):
firstfile = file
lastfile = file
nTrees = 0
nHalos = 0
nTreeHalos = numpy.array([],dtype=numpy.int32)
output_Halos = numpy.array([],dtype=struct_lgalinput)
output_HaloIDs = numpy.array([],dtype=struct_lgaldbidsinput)
ifile = file
filename = folder+'/trees_'+"%03d"%(lastsnap)+'.'+"%d"%(ifile)
f = open(filename,"rb")
this_nTrees = numpy.fromfile(f,numpy.int32,1)[0]
nTrees += this_nTrees
this_nHalos = numpy.fromfile(f,numpy.int32,1)[0]
nHalos += this_nHalos
if(verbose):
print "File ", ifile," nHalos = ",this_nHalos
nTreeHalos = numpy.fromfile(f,numpy.int32,this_nTrees)
output_Halos = numpy.fromfile(f,struct_lgalinput,this_nHalos)
f.close()
filename = folder+'/tree_dbids_'+"%03d"%(lastsnap)+'.'+"%d"%(ifile)
f = open(filename,"rb")
output_HaloIDs = numpy.fromfile(f,struct_lgaldbidsinput,this_nHalos)
f.close()
return (nTrees,nHalos,nTreeHalos,output_Halos,output_HaloIDs)
def __parse_data(self):
print 'Reading ', self._filepath
f = open(self._filepath, 'rb')
samp_per_segment = 64
bytes_per_sample = 2
channels = 2
tcd_dtype= 'int16'
f_size = os.path.getsize(self._filepath)
segments = f_size / ( samp_per_segment * bytes_per_sample * channels )
self._progress_bar.setMinimum(0)
self._progress_bar.setMaximum(segments)
self._value = 0
self._progress_bar.setValue(self._value)
chan1 = numpy.array([], dtype=tcd_dtype)
chan2 = numpy.array([], dtype=tcd_dtype)
data = numpy.zeros((samp_per_segment), dtype=tcd_dtype)
for seg in xrange(segments):
self._value = self._value + 1
self._progress_bar.setValue(self._value)
data = numpy.fromfile(f, dtype=tcd_dtype, count=samp_per_segment)
chan1 = numpy.concatenate((chan1, data.copy()) )
data = numpy.fromfile(f, dtype=tcd_dtype, count=samp_per_segment)
chan2 = numpy.concatenate((chan2, data.copy()) )
f.close()
chan1 = chan1.astype(float) / 2.0**11 * self._prf/2.0 *154000.0 / self._doppler_freq_1/10**3
chan2 = chan2.astype(float) / 2.0**11 * self._prf/2.0 *154000.0 / self._doppler_freq_2/ 10**3
self.chan1 = chan1
self.chan2 = chan2
def openDATfile(filename,ftype,srate=25000):
fh = open(filename,'r')
fh.seek(0)
if ftype == 'amp':
data = np.fromfile(fh, dtype=np.int16)
fh.close()
data = np.double(data)
data *= 0.195 # according the Intan, the output should be multiplied by 0.195 to be converted to micro-volts
elif ftype == 'adc':
data = np.fromfile(fh, dtype=np.uint16)
fh.close()
data = np.double(data)
data *= 0.000050354 # according the Intan, the output should be multiplied by 0.195 to be converted to micro-volts
data -= np.mean(data)
elif ftype == 'aux':
data = np.fromfile(fh, dtype=np.uint16)
fh.close()
data = np.double(data)
data *= 0.0000748 # according the Intan, the output should be multiplied by 0.195 to be converted to micro-volts
elif ftype == 'time':
data = np.fromfile(fh, dtype=np.int32)
fh.close()
data = np.double(data)
data /= srate # according the Intan, the output should be multiplied by 0.195 to be converted to micro-volts
return data
def load_matrix_csr(path, verbose=False):
t_start = time.time()
data = np.fromfile(
open(
os.path.join(
path,
"bigrams.data.bin")),
dtype=np.float32)
col_ind = np.fromfile(
open(
os.path.join(
path,
"bigrams.col_ind.bin")),
dtype=np.int64)
row_ptr = np.fromfile(
open(
os.path.join(
path,
"bigrams.row_ptr.bin")),
dtype=np.int64)
dim = row_ptr.shape[0] - 1
cooccurrence = scipy.sparse.csr_matrix(
(data, col_ind, row_ptr), shape=(
dim, dim), dtype=np.float32)
t_end = time.time()
if verbose:
print("Matrix loaded in {0:0.2f} sec".format(t_end - t_start))
print_stats(cooccurrence)
return cooccurrence
def read_from_file(self, filename):
'''
Read data from file. Sets the instance variables
self.raw_velocity and self.kmsrho8
Parameters:
* filename (string): the file to read from.
Returns:
Nothing
'''
print_msg('Reading velocity file: %s...' % filename)
self.filename = filename
#Read raw data from velocity file
f = open(filename, 'rb')
temp_mesh = np.fromfile(f, count=3, dtype='int32')
self.mesh_x, self.mesh_y, self.mesh_z = temp_mesh
self.raw_velocity = np.fromfile(f, dtype='float32').astype('float64')
f.close()
self.raw_velocity = self.raw_velocity.reshape((3, self.mesh_x, self.mesh_y, self.mesh_z), order='F')
#Store the redshift from the filename
try:
import os.path
name = os.path.split(filename)[1]
self.z = float(name.split('v_')[0])
except:
print_msg('Could not determine redshift from file name')
self.z = -1
#Convert to kms/s*(rho/8)
self.kmsrho8 = self.raw_velocity*conv.velconvert(z = self.z)
print_msg('...done')
def read(digits, dataset = "training", path = "."):
"""
Python function for importing the MNIST data set.
"""
if dataset is "training":
fname_img = os.path.join(path, 'MNIST_data/train-images-idx3-ubyte')
fname_lbl = os.path.join(path, 'MNIST_data/train-labels-idx1-ubyte')
elif dataset is "testing":
fname_img = os.path.join(path, 'MNIST_data/t10k-images-idx3-ubyte')
fname_lbl = os.path.join(path, 'MNIST_data/t10k-labels-idx1-ubyte')
else:
raise ValueError, "dataset must be 'testing' or 'training'"
flbl = open(fname_lbl, 'rb')
magic_nr, size = struct.unpack(">II", flbl.read(8))
lbl = np.fromfile(flbl, dtype=np.int8)
flbl.close()
fimg = open(fname_img, 'rb')
magic_nr, size, rows, cols = struct.unpack(">IIII", fimg.read(16))
img = np.fromfile(fimg, dtype=np.uint8).reshape(len(lbl), rows, cols)
fimg.close()
ind = [ k for k in xrange(size) if lbl[k] in digits ]
images = np.zeros((len(ind),rows,cols))
labels = np.zeros((len(ind),1))
for i in xrange(len(ind)):
images[i, :] = img[ ind[i],:,:]
labels[i] = lbl[ind[i]]
return images, labels
def test_masked_gauss(self):
data = numpy.ones((50, 10))
data[:, 5:] = 2
lons = numpy.fromfunction(lambda y, x: 3 + x, (50, 10))
lats = numpy.fromfunction(lambda y, x: 75 - y, (50, 10))
swath_def = geometry.SwathDefinition(lons=lons, lats=lats)
mask = numpy.ones((50, 10))
mask[:, :5] = 0
masked_data = numpy.ma.array(data, mask=mask)
res = kd_tree.resample_gauss(swath_def, masked_data.ravel(),
self.area_def, 50000, 25000, segments=1)
expected_mask = numpy.fromfile(os.path.join(os.path.dirname(__file__),
'test_files',
'mask_test_mask.dat'),
sep=' ').reshape((800, 800))
expected_data = numpy.fromfile(os.path.join(os.path.dirname(__file__),
'test_files',
'mask_test_data.dat'),
sep=' ').reshape((800, 800))
expected = expected_data.sum()
cross_sum = res.data.sum()
self.assertTrue(numpy.array_equal(expected_mask, res.mask),
msg='Gauss resampling of swath mask failed')
self.assertAlmostEqual(cross_sum, expected, places=3,
msg='Gauss resampling of swath masked data failed')
def _array_from_file(infile, dtype, count, sep):
"""Create a numpy array from a file or a file-like object."""
if isfile(infile):
global CHUNKED_FROMFILE
if CHUNKED_FROMFILE is None:
if (sys.platform == 'darwin' and
LooseVersion(platform.mac_ver()[0]) < LooseVersion('10.9')):
CHUNKED_FROMFILE = True
else:
CHUNKED_FROMFILE = False
if CHUNKED_FROMFILE:
chunk_size = int(1024 ** 3 / dtype.itemsize) # 1Gb to be safe
if count < chunk_size:
return np.fromfile(infile, dtype=dtype, count=count, sep=sep)
else:
array = np.empty(count, dtype=dtype)
for beg in range(0, count, chunk_size):
end = min(count, beg + chunk_size)
array[beg:end] = np.fromfile(infile, dtype=dtype, count=end - beg, sep=sep)
return array
else:
return np.fromfile(infile, dtype=dtype, count=count, sep=sep)
else:
# treat as file-like object with "read" method; this includes gzip file
# objects, because numpy.fromfile just reads the compressed bytes from
# their underlying file object, instead of the decompressed bytes
read_size = np.dtype(dtype).itemsize * count
s = infile.read(read_size)
return np.fromstring(s, dtype=dtype, count=count, sep=sep)
def _RunWebRtcApmVad(self, wav_file_path):
# Create temporary output path.
tmp_path = tempfile.mkdtemp()
output_file_path_probs = os.path.join(
tmp_path, os.path.split(wav_file_path)[1] + '_vad_probs.tmp')
output_file_path_rms = os.path.join(
tmp_path, os.path.split(wav_file_path)[1] + '_vad_rms.tmp')
# Call WebRTC VAD.
try:
subprocess.call([
self._VAD_WEBRTC_APM_PATH,
'-i', wav_file_path,
'-o_probs', output_file_path_probs,
'-o_rms', output_file_path_rms
], cwd=self._VAD_WEBRTC_PATH)
# Parse annotations.
self._apm_vad_probs = np.fromfile(output_file_path_probs, np.double)
self._apm_vad_rms = np.fromfile(output_file_path_rms, np.double)
assert len(self._apm_vad_rms) == len(self._apm_vad_probs)
except Exception as e:
logging.error('Error while running the WebRTC APM VAD (' +
e.message + ')')
finally:
if os.path.exists(tmp_path):
shutil.rmtree(tmp_path)
def parseDataFile(self, uLong, precision, encoding, dataFile):
assert uLong in ['uint32', 'uint64']
assert precision in ['single', 'double']
assert encoding in ['BigEndian', 'LittleEndian']
require_numpy()
fd = file(dataFile, 'rb')
byteorder = {'LittleEndian': '<', 'BigEndian': '>'}[encoding]
unsignedLongTypeString = {'uint32': 'u4', 'uint64': 'u8'}[uLong]
realTypeString = {'single': 'f4', 'double': 'f8'}[precision]
ulongDType = numpy.dtype(byteorder + unsignedLongTypeString)
floatDType = numpy.dtype(byteorder + realTypeString)
independentGeometry = []
for independentVariable in self.independentVariables:
size = numpy.fromfile(fd, dtype=ulongDType, count=1)
independentGeometry.append(size)
assert size == independentVariable['length']
a = numpy.fromfile(fd, dtype=floatDType, count=size)
independentVariable['array'] = a
if len(independentGeometry) == 0:
independentGeometry.append(1)
for dependentVariable in self.dependentVariables:
size = numpy.fromfile(fd, dtype=ulongDType, count=1)
a = numpy.fromfile(fd, dtype=floatDType, count=size)
assert a.size == size, "Data file %s has incorrect size. Variable '%s' wasn't written completely." % (dataFile, dependentVariable['name'])
dependentVariable['array'] = a.reshape(*independentGeometry)
def read_record(self, dtype='b1'):
"""
Read and return a record of numpy type dtype from the current file.
If the record is a single value, it is returned.
Otherwise, a numpy.ndarray is returned.
dtype is the data type to read (Python type or numpy dtype or string
identifier).
"""
if self._mode != 'r' and self._mode != 'rb':
raise FortranIOException('Not in read mode')
dtype = np.dtype(dtype)
nbytes = self._read_control()
nitems = nbytes // dtype.itemsize
if nbytes % dtype.itemsize != 0:
raise FortranIOException('Record size not valid for data type')
if nitems > 1:
data = np.fromfile(self._file, dtype, nitems)
else:
data = np.fromfile(self._file, dtype, nitems)[0]
nbytes2 = self._read_control()
if nbytes != nbytes2:
raise FortranIOException('Record head and tail mismatch')
return data
def load_ply(fn):
f = open(fn, 'rb')
prop_to_dtype = { 'float' : np.float32, 'int' : np.int32, 'uchar' : np.uint8 }
header = []
while True:
s = f.readline().split()
header.append(s)
if s[0] == 'end_header':
break
it = iter(header)
s = it.next()
elements = {}
while True:
if s[0] == 'end_header':
break
if s[0] == 'element':
el_name, el_len = s[1], int(s[2])
el_props = []
s = it.next()
while s[0] == 'property':
el_props.append( s )
s = it.next()
if el_name == 'face':
el_type = np.dtype( [('count', np.uint8), ('idx', np.int32, 3)] )
elements[el_name] = np.fromfile(f, el_type, el_len)['idx'].copy()
else:
el_type = np.dtype( [(name, np.dtype(prop_to_dtype[tp])) for _, tp, name in el_props] )
elements[el_name] = np.fromfile(f, el_type, el_len)
continue
s = it.next()
return elements
def __init__(self, filename, verbose = False):
super(SpeReader, self).__init__(filename, verbose = verbose)
# open the file & read the header
self.header_size = 4100
self.fileptr = open(filename, "rb")
self.fileptr.seek(42)
self.image_width = int(numpy.fromfile(self.fileptr, numpy.uint16, 1)[0])
self.fileptr.seek(656)
self.image_height = int(numpy.fromfile(self.fileptr, numpy.uint16, 1)[0])
self.fileptr.seek(1446)
self.number_frames = int(numpy.fromfile(self.fileptr, numpy.uint32, 1)[0])
self.fileptr.seek(108)
image_mode = int(numpy.fromfile(self.fileptr, numpy.uint16, 1)[0])
if (image_mode == 0):
self.image_size = 4 * self.image_width * self.image_height
self.image_mode = numpy.float32
elif (image_mode == 1):
self.image_size = 4 * self.image_width * self.image_height
self.image_mode = numpy.uint32
elif (image_mode == 2):
self.image_size = 2 * self.image_width * self.image_height
self.image_mode = numpy.int16
elif (image_mode == 3):
self.image_size = 2 * self.image_width * self.image_height
self.image_mode = numpy.uint16
else:
print("unrecognized spe image format: ", image_mode)
def parseData(dataset='testing', path='.'):
'''
parseData - Parses a file into matrices
Input - the name of file to be parsed
Output - The data in matrix representation
'''
if dataset == 'training':
image_file = os.path.join(path, 'train-images-idx3-ubyte')
label_file = os.path.join(path, 'train-labels-idx1-ubyte')
elif dataset == 'testing':
image_file = os.path.join(path, 't10k-images-idx3-ubyte')
label_file = os.path.join(path, 't10k-labels-idx1-ubyte')
else:
raise(ValueError, "'dataset' must be in testing or 'training'")
# get the matrix for image data
f_img = open(image_file, 'rb')
magic_nr, size = struct.unpack(">II", f_img.read(8)) # parse the magic number, & size of dataset
dim_x, dim_y = struct.unpack(">II", f_img.read(8)) # get the dimensions of each handwritten num
X = np.fromfile(f_img, dtype=np.dtype('B'))
X = X.reshape(size, dim_x * dim_y)
# get the matrix for label data
f_lbl = open(label_file, 'rb')
magic_nr, size = struct.unpack(">II", f_lbl.read(8)) # only magic # and size of dataset
y = np.fromfile(f_lbl, dtype=np.dtype('B'))
#X[X > 1] = 1
return X, y
def __init__(self, filename, xml):
DataReader.__init__(self, filename, xml)
# Open the file & read the header.
self.header_size = 4100
self.fileptr = open(filename, "rb")
# FIXME: Should check that these match the XML file.
self.fileptr.seek(42)
self.image_width = int(numpy.fromfile(self.fileptr, numpy.uint16, 1)[0])
self.fileptr.seek(656)
self.image_height = int(numpy.fromfile(self.fileptr, numpy.uint16, 1)[0])
self.fileptr.seek(1446)
self.number_frames = int(numpy.fromfile(self.fileptr, numpy.uint32, 1)[0])
self.fileptr.seek(108)
image_mode = int(numpy.fromfile(self.fileptr, numpy.uint16, 1)[0])
if (image_mode == 0):
self.image_size = 4 * self.image_width * self.image_height
self.image_mode = numpy.float32
elif (image_mode == 1):
self.image_size = 4 * self.image_width * self.image_height
self.image_mode = numpy.uint32
elif (image_mode == 2):
self.image_size = 2 * self.image_width * self.image_height
self.image_mode = numpy.int16
elif (image_mode == 3):
self.image_size = 2 * self.image_width * self.image_height
self.image_mode = numpy.uint16
else:
print "unrecognized spe image format: ", image_mode
def getPressureData(fullFileName):
itemsize = os.path.getsize(fullFileName)/8/2
fp = open(fullFileName, 'r')
time = np.fromfile(fp, dtype='Float64', count=itemsize);
pressure = np.fromfile(fp, dtype='Float64', count=itemsize);
fp.close()
return (time, pressure)
def readFlowFile(file_name,flip=False):
data2D=None
with open(file_name,'rb') as f:
magic = np.fromfile(f, np.float32, count=1)
if 202021.25 != magic:
print 'Magic number incorrect. Invalid .flo file'
else:
w = np.fromfile(f, np.int32, count=1)
h = np.fromfile(f, np.int32, count=1)
if w.size==0 or h.size==0:
# print type(w),type(h),w,h
data2D=None;
else:
# print (w, h)
data = np.fromfile(f, np.float32, count=2*w*h)
# Reshape data into 3D array (columns, rows, bands)
# if flip is True:
# data2D = np.resize(data, (w, h, 2))
# data2D = data2D.
# data2D = np.reshape(data, (h, w, 2))
# # ,order='F')
# else:
data2D = np.reshape(data, (h, w, 2))
# print data2D.shape
return data2D
def sorted_indexes(filename,data_type=np.uint32):
if os.path.exists(filename):
curfilename = filename
elif os.path.exists(filename+".0"):
curfilename = filename+".0"
else:
print "file not found:", filename
sys.exit()
f=open(curfilename,'rb')
number_of_files=np.fromfile(f,dtype=np.uint32,count=1)[0]
dims=np.fromfile(f,dtype=np.uint32,count=1)[0]
dims3=dims**3
total_size=np.fromfile(f,dtype=data_type,count=dims3)
total_array=[]
for j in range(dims3):
total_array.append(np.empty(total_size[j],dtype=data_type))
f.close()
total_array=np.array(total_array)
offset=np.zeros(dims3,dtype=data_type)
for i in range(number_of_files):
curfilename=filename+'.'+str(i)
f=open(curfilename,'rb')
f.seek(4*(2+dims3),os.SEEK_CUR)
for j in range(dims3):
size=np.fromfile(f,dtype=data_type,count=1)[0]
array=np.fromfile(f,dtype=data_type,count=size)
total_array[j][offset[j]:offset[j]+size]=array
offset[j]+=size
f.close()
return total_array
def readFLO(path):
f = open(path, 'rb')
# Read magic number ("PIEH" in ASCII = float 202021.25)
magic = np.fromfile(f, np.float32, count=1)
if magic != 202021.25:
raise Exception('Invalid .flo file')
# Read width
f.seek(4)
w = np.fromfile(f, np.int32, count=1)
# Read height
f.seek(8)
h = np.fromfile(f, np.int32, count=1)
# Read (u,v) coordinates
f.seek(12)
data = np.fromfile(f, np.float32, count=w*h*2)
# Close file (.flo)
f.close()
# Reshape data into 3D array (columns, rows, bands)
dataM = np.resize(data, (h, w, 2))
# Extract u and v coordinates
u = dataM[:,:,0]
v = dataM[:,:,1]
return w,h,u,v
def read(dataset = "training", path = "."):
"""
Python function for importing the MNIST data set. It returns an iterator
of 2-tuples with the first element being the label and the second element
being a numpy.uint8 2D array of pixel data for the given image.
"""
if dataset is "training":
fname_img = os.path.join(path, 'train-images-idx3-ubyte')
fname_lbl = os.path.join(path, 'train-labels-idx1-ubyte')
elif dataset is "testing":
fname_img = os.path.join(path, 't10k-images-idx3-ubyte')
fname_lbl = os.path.join(path, 't10k-labels-idx1-ubyte')
else:
raise ValueError, "dataset must be 'testing' or 'training'"
# Load everything in some numpy arrays
with open(fname_lbl, 'rb') as flbl:
magic, num = struct.unpack(">II", flbl.read(8))
lbl = np.fromfile(flbl, dtype=np.int8)
with open(fname_img, 'rb') as fimg:
magic, num, rows, cols = struct.unpack(">IIII", fimg.read(16))
img = np.fromfile(fimg, dtype=np.uint8).reshape(len(lbl), rows, cols)
get_img = lambda idx: (lbl[idx], img[idx])
# Create an iterator which returns each image in turn
for i in xrange(len(lbl)):
yield get_img(i)
def my_fread(*x, **y):
return np.fromfile(*x, **y)[0]
def read_geometry(filepath, read_metadata=False, read_stamp=False):
"""Read a triangular format Freesurfer surface mesh.
Parameters
----------
filepath : str
Path to surface file.
read_metadata : bool, optional
If True, read and return metadata as key-value pairs.
Valid keys:
* 'head' : array of int
* 'valid' : str
* 'filename' : str
* 'volume' : array of int, shape (3,)
* 'voxelsize' : array of float, shape (3,)
* 'xras' : array of float, shape (3,)
* 'yras' : array of float, shape (3,)
* 'zras' : array of float, shape (3,)
* 'cras' : array of float, shape (3,)
read_stamp : bool, optional
Return the comment from the file
Returns
-------
coords : numpy array
nvtx x 3 array of vertex (x, y, z) coordinates.
faces : numpy array
nfaces x 3 array of defining mesh triangles.
volume_info : OrderedDict
Returned only if `read_metadata` is True. Key-value pairs found in the
geometry file.
create_stamp : str
Returned only if `read_stamp` is True. The comment added by the
program that saved the file.
"""
volume_info = OrderedDict()
TRIANGLE_MAGIC = 16777214
QUAD_MAGIC = 16777215
NEW_QUAD_MAGIC = 16777213
with open(filepath, "rb") as fobj:
magic = _fread3(fobj)
if magic in (QUAD_MAGIC, NEW_QUAD_MAGIC): # Quad file
nvert = _fread3(fobj)
nquad = _fread3(fobj)
(fmt, div) = (">i2", 100.) if magic == QUAD_MAGIC else (">f4", 1.)
coords = np.fromfile(fobj, fmt, nvert * 3).astype(np.float64) / div
coords = coords.reshape(-1, 3)
quads = _fread3_many(fobj, nquad * 4)
quads = quads.reshape(nquad, 4)
#
# Face splitting follows
#
faces = np.zeros((2 * nquad, 3), dtype=int)
nface = 0
for quad in quads:
if (quad[0] % 2) == 0:
faces[nface] = quad[0], quad[1], quad[3]
nface += 1
faces[nface] = quad[2], quad[3], quad[1]
nface += 1
else:
faces[nface] = quad[0], quad[1], quad[2]
nface += 1
faces[nface] = quad[0], quad[2], quad[3]
nface += 1
elif magic == TRIANGLE_MAGIC: # Triangle file
create_stamp = fobj.readline().rstrip(b'\n').decode('utf-8')
fobj.readline()
vnum = np.fromfile(fobj, ">i4", 1)[0]
fnum = np.fromfile(fobj, ">i4", 1)[0]
coords = np.fromfile(fobj, ">f4", vnum * 3).reshape(vnum, 3)
faces = np.fromfile(fobj, ">i4", fnum * 3).reshape(fnum, 3)
if read_metadata:
volume_info = _read_volume_info(fobj)
else:
raise ValueError("File does not appear to be a Freesurfer surface")
coords = coords.astype(np.float64) # XXX: due to mayavi bug on mac 32bits
ret = (coords, faces)
if read_metadata:
if len(volume_info) == 0:
warnings.warn('No volume information contained in the file')
ret += (volume_info, )
if read_stamp:
ret += (create_stamp, )
return ret
def load_velo_scan(file):
'''Load and parse a velodyne binary file'''
scan = np.fromfile(file, dtype=np.float32)
return scan.reshape((-1, 4))
def readsu(filename,nx,ny,nz,timeslice): ff=open(filename,mode='rb') ff.seek(8*timeslice*nx*ny*nz) field = np.fromfile(ff,dtype='float64',count=nx*ny*nz).reshape(field,(nx,ny,nz),order='F') ff.close() return field
def __init__(self,
basedir,
snapnum,
long_ids=False,
swap=False,
SFR=False,
read_IDs=True,
prefix='/groups_'):
if long_ids: format = np.uint64
else: format = np.uint32
exts = ('000' + str(snapnum))[-3:]
################# READ TAB FILES #################
fnb, skip, Final = 0, 0, False
dt1 = np.dtype((np.float32, 3))
dt2 = np.dtype((np.float32, 6))
prefix = basedir + prefix + exts + "/group_tab_" + exts + "."
while not (Final):
f = open(prefix + str(fnb), 'rb')
self.Ngroups = np.fromfile(f, dtype=np.int32, count=1)[0]
self.TotNgroups = np.fromfile(f, dtype=np.int32, count=1)[0]
self.Nids = np.fromfile(f, dtype=np.int32, count=1)[0]
self.TotNids = np.fromfile(f, dtype=np.uint64, count=1)[0]
self.Nfiles = np.fromfile(f, dtype=np.uint32, count=1)[0]
TNG, NG = self.TotNgroups, self.Ngroups
if fnb == 0:
self.GroupLen = np.empty(TNG, dtype=np.int32)
self.GroupOffset = np.empty(TNG, dtype=np.int32)
self.GroupMass = np.empty(TNG, dtype=np.float32)
self.GroupPos = np.empty(TNG, dtype=dt1)
self.GroupVel = np.empty(TNG, dtype=dt1)
self.GroupTLen = np.empty(TNG, dtype=dt2)
self.GroupTMass = np.empty(TNG, dtype=dt2)
if SFR: self.GroupSFR = np.empty(TNG, dtype=np.float32)
if NG > 0:
locs = slice(skip, skip + NG)
self.GroupLen[locs] = np.fromfile(f, dtype=np.int32, count=NG)
self.GroupOffset[locs] = np.fromfile(f,
dtype=np.int32,
count=NG)
self.GroupMass[locs] = np.fromfile(f,
dtype=np.float32,
count=NG)
self.GroupPos[locs] = np.fromfile(f, dtype=dt1, count=NG)
self.GroupVel[locs] = np.fromfile(f, dtype=dt1, count=NG)
self.GroupTLen[locs] = np.fromfile(f, dtype=dt2, count=NG)
self.GroupTMass[locs] = np.fromfile(f, dtype=dt2, count=NG)
if SFR:
self.GroupSFR[locs] = np.fromfile(f,
dtype=np.float32,
count=NG)
skip += NG
if swap:
self.GroupLen.byteswap(True)
self.GroupOffset.byteswap(True)
self.GroupMass.byteswap(True)
self.GroupPos.byteswap(True)
self.GroupVel.byteswap(True)
self.GroupTLen.byteswap(True)
self.GroupTMass.byteswap(True)
if SFR: self.GroupSFR.byteswap(True)
curpos = f.tell()
f.seek(0, os.SEEK_END)
if curpos != f.tell():
raise Exception(
"Warning: finished reading before EOF for tab file", fnb)
f.close()
fnb += 1
if fnb == self.Nfiles: Final = True
################# READ IDS FILES #################
if read_IDs:
fnb, skip = 0, 0
Final = False
while not (Final):
fname = basedir + "/groups_" + exts + "/group_ids_" + exts + "." + str(
fnb)
f = open(fname, 'rb')
Ngroups = np.fromfile(f, dtype=np.uint32, count=1)[0]
TotNgroups = np.fromfile(f, dtype=np.uint32, count=1)[0]
Nids = np.fromfile(f, dtype=np.uint32, count=1)[0]
TotNids = np.fromfile(f, dtype=np.uint64, count=1)[0]
Nfiles = np.fromfile(f, dtype=np.uint32, count=1)[0]
Send_offset = np.fromfile(f, dtype=np.uint32, count=1)[0]
if fnb == 0:
self.GroupIDs = np.zeros(dtype=format, shape=TotNids)
if Ngroups > 0:
if long_ids:
IDs = np.fromfile(f, dtype=np.uint64, count=Nids)
else:
IDs = np.fromfile(f, dtype=np.uint32, count=Nids)
if swap:
IDs = IDs.byteswap(True)
self.GroupIDs[skip:skip + Nids] = IDs[:]
skip += Nids
curpos = f.tell()
f.seek(0, os.SEEK_END)
if curpos != f.tell():
raise Exception(
"Warning: finished reading before EOF for IDs file",
fnb)
f.close()
fnb += 1
if fnb == Nfiles: Final = True
def get_sensor_data(self, query):
idx = query
read_test_image = False
if isinstance(query, dict):
assert "lidar" in query
idx = query["lidar"]["idx"]
read_test_image = "cam" in query
info = self._nusc_infos[idx]
res = {
"lidar": {
"type": "lidar",
"points": None,
},
"metadata": {
"token": info["token"],
"scene_token": info["scene_token"],
},
}
lidar_path = Path(info['lidar_path'])
points = np.fromfile(str(lidar_path), dtype=np.float32,
count=-1).reshape([-1, 5])
points[:, 3] /= 255
points[:, 4] = 0
sweep_points_list = [points]
ts = info["timestamp"] / 1e6
for sweep in info["sweeps"]:
points_sweep = np.fromfile(str(sweep["lidar_path"]),
dtype=np.float32,
count=-1).reshape([-1, 5])
sweep_ts = sweep["timestamp"] / 1e6
points_sweep[:, 3] /= 255
points_sweep[:, :3] = points_sweep[:, :3] @ sweep[
"sweep2lidar_rotation"].T
points_sweep[:, :3] += sweep["sweep2lidar_translation"]
points_sweep[:, 4] = ts - sweep_ts
sweep_points_list.append(points_sweep)
points = np.concatenate(sweep_points_list, axis=0)[:, [0, 1, 2, 4]]
if read_test_image:
if Path(info["cam_front_path"]).exists():
with open(str(info["cam_front_path"]), 'rb') as f:
image_str = f.read()
else:
image_str = None
res["cam"] = {
"type": "camera",
"data": image_str,
"datatype": Path(info["cam_front_path"]).suffix[1:],
}
res["lidar"]["points"] = points
if 'gt_boxes' in info:
mask = info["num_lidar_pts"] > 0
gt_boxes = info["gt_boxes"][mask]
if self._with_velocity:
gt_velocity = info["gt_velocity"][mask]
nan_mask = np.isnan(gt_velocity[:, 0])
gt_velocity[nan_mask] = [0.0, 0.0]
gt_boxes = np.concatenate([gt_boxes, gt_velocity], axis=-1)
res["lidar"]["annotations"] = {
'boxes': gt_boxes,
'names': info["gt_names"][mask],
}
return res
rate_drop_lstm = 0.15 + np.random.rand() * 0.25
rate_drop_dense = 0.15 + np.random.rand() * 0.25
act = 'relu'
re_weight = False # whether to re-weight classes to fit the 17.5% share in test set
# NOT FOR THE PROJECT BECAUSE NOT USING TEST
STAMP = 'lstm_%d_%d_%.2f_%.2f'%(num_lstm, num_dense, rate_drop_lstm, \
rate_drop_dense)
########################################
## uploads files
########################################
# upload premade embedding
embedded = np.fromfile(EMBEDDING_FILE, dtype=EMBEDDED_TYPE).reshape(
(COLUMNS_PREPROCESS, -1)).transpose()
# store pair id and duplicate
pair_id = list(map(int, embedded[:, -1]))
question_id = list(map(int, embedded[:, -2]))
labels = np.array(list(map(int, embedded[:, -3])))
# removes unnecessary column from embedding
embedded = embedded[:, :-3]
data_1 = np.fromfile(Q1_FILE, dtype=EMBEDDED_TYPE).reshape(
(-1, MAX_SEQUENCE_LENGTH))
data_2 = np.fromfile(Q2_FILE, dtype=EMBEDDED_TYPE).reshape(
(-1, MAX_SEQUENCE_LENGTH))
########################################
## sample train/validation data
print 'Reading data in ' + float_type + ' format from ' + infile
indata.read(1) # the '_' symbol which starts VTK appended data
bytecount = struct.unpack('i', indata.read(4))[0]
#define the expected structure
if 'little' in endianness.lower():
float_format = '<'
elif 'big' in endianness.lower():
float_format = '>'
else:
raise Exception("Didn't understand endianness of input: " + endiannness)
if '32' in float_type:
float_format += 'f'
num_count = bytecount / 4
elif '64' in float_type:
float_format += 'd'
num_count = bytecount / 8
else:
raise Exception("Didn't understand float format for points: " + float_type)
np_data = np.fromfile(indata, dtype=np.dtype(float_format),
count=num_count).reshape((-1, 3))
indata.close()
np.savetxt(outfile, np_data, fmt='%16.6f')
import matplotlib
import matplotlib.pyplot as plt
import matplotlib.animation as animation
from scipy import signal
import librosa
import dft_freq_est
import attack_finder
from common import normalize
def cat_to_cols(x,y):
return np.concatenate((x[:,None],y[:,None]),axis=1)
do_plot=False
sr=16000
x=np.fromfile('/tmp/snd.f64')
x=normalize(x)
N_W=1024
N_FFT=2048
N_H=2048
n_max=((len(x)-N_W)//N_H)*N_H
peak_thresh=-140
# intialize the sinusoidal analyser
pa=dft_freq_est.peak_analyzer(
N_FFT,
N_W,
N_H)
def __call__(self, results):
"""Call functions to load image and get image meta information.
Args:
results (dict): Result dict from :obj:`mmseg.CustomDataset`.
Returns:
dict: The dict contains loaded image and meta information.
"""
if results.get('img_prefix') is not None:
filename = osp.join(results['img_prefix'],
results['img_info']['filename'])
else:
filename = results['img_info']['filename']
hdr = dict()
with open(filename) as f:
for line in f.readlines():
if '=' not in line:
continue
else:
key, value = line.split('=')
key = key.strip()
value = value.strip()
hdr[key] = value
assert hdr[
'file type'] == 'ENVI Standard', 'Require ENVI data: file type = ENVI Standard'
assert hdr['byte order'] == '0', 'Require ENVI data: byte order = 0'
assert hdr['x start'] == '0', 'Require ENVI data: x start = 0'
assert hdr['y start'] == '0', 'Require ENVI data: y start = 0'
assert hdr['interleave'].lower(
) == 'bsq', 'Require ENVI data: interleave = bsq'
assert int(hdr['data type']) <= len(
self.ENVI_data_type) and self.ENVI_data_type[int(
hdr['data type'])] != None
data_type = int(hdr['data type'])
header_offset = int(hdr['header offset'])
height = int(hdr['lines'])
width = int(hdr['samples'])
bands = int(hdr['bands'])
if hdr['interleave'].lower() == 'bsq':
img_bytes = np.fromfile(filename.replace('.hdr', '.raw'),
dtype=self.ENVI_data_type[data_type],
offset=header_offset)
img_bytes = img_bytes.reshape((bands, height, width))
img_bytes = img_bytes[self.channel_select, :, :]
if self.dataset_name == 'cholangiocarcinoma':
img_bytes = img_bytes[:, ::-1, :]
img_bytes = np.transpose(img_bytes, (1, 2, 0))
else:
img_bytes = np.zeros((height, width, bands),
dtype=self.ENVI_data_type[data_type])
pass
if self.to_float32:
img_bytes = img_bytes.astype(np.float32)
if self.normalization:
img_bytes -= self.mean[..., self.channel_select]
img_bytes /= self.std[..., self.channel_select]
############################################3
# img_bytes *= 16
# img_bytes += 128
# img_bytes = img_bytes.astype(np.uint8)
# img_bytes = img_bytes.astype(np.float32)
# img_bytes -= 128
# img_bytes /= 16
##############################################
if self.median_blur:
for band in range(img_bytes.shape[0]):
img_bytes[band, :, :] = cv2.medianBlur(img_bytes[band, :, :],
ksize=3)
results['filename'] = filename.replace('.hdr', '.png')
results['ori_filename'] = results['img_info']['filename'].replace(
'.hdr', '.png')
results['img'] = img_bytes
results['img_shape'] = img_bytes.shape
results['ori_shape'] = img_bytes.shape
# Set initial values for default meta_keys
results['pad_shape'] = img_bytes.shape
results['scale_factor'] = 1.0
results['channel_select'] = self.channel_select
results['channel_to_show'] = self.channel_to_show
num_channels = 1 if len(img_bytes.shape) < 3 else img_bytes.shape[2]
mean = np.ones(num_channels, dtype=np.float32) * 128
std = np.ones(num_channels, dtype=np.float32) * 16
results['img_norm_cfg'] = dict(mean=mean, std=std, to_rgb=False)
return results
def _is_valid(cls, filename, *args, **kwargs):
"""
Defined for the NMSU file naming scheme.
This could differ for other formats.
"""
f = str(filename)
prefix, suffix = filename_pattern["particle_data"]
if not os.path.isfile(f):
return False
if not f.endswith(suffix):
return False
if "s0" not in f:
# ATOMIC.DAT, for instance, passes the other tests, but then dies
# during _find_files because it can't be split.
return False
with open(f, "rb") as fh:
try:
amr_prefix, amr_suffix = filename_pattern["amr"]
possibles = glob.glob(os.path.dirname(os.path.abspath(f)) + "/*")
for possible in possibles:
if possible.endswith(amr_suffix):
if os.path.basename(possible).startswith(amr_prefix):
return False
except Exception:
pass
try:
seek = 4
fh.seek(seek)
headerstr = np.fromfile(fh, count=1, dtype=(str, 45)) # NOQA
aexpn = np.fromfile(fh, count=1, dtype=">f4") # NOQA
aexp0 = np.fromfile(fh, count=1, dtype=">f4") # NOQA
amplt = np.fromfile(fh, count=1, dtype=">f4") # NOQA
astep = np.fromfile(fh, count=1, dtype=">f4") # NOQA
istep = np.fromfile(fh, count=1, dtype=">i4") # NOQA
partw = np.fromfile(fh, count=1, dtype=">f4") # NOQA
tintg = np.fromfile(fh, count=1, dtype=">f4") # NOQA
ekin = np.fromfile(fh, count=1, dtype=">f4") # NOQA
ekin1 = np.fromfile(fh, count=1, dtype=">f4") # NOQA
ekin2 = np.fromfile(fh, count=1, dtype=">f4") # NOQA
au0 = np.fromfile(fh, count=1, dtype=">f4") # NOQA
aeu0 = np.fromfile(fh, count=1, dtype=">f4") # NOQA
nrowc = np.fromfile(fh, count=1, dtype=">i4") # NOQA
ngridc = np.fromfile(fh, count=1, dtype=">i4") # NOQA
nspecs = np.fromfile(fh, count=1, dtype=">i4") # NOQA
nseed = np.fromfile(fh, count=1, dtype=">i4") # NOQA
Om0 = np.fromfile(fh, count=1, dtype=">f4") # NOQA
Oml0 = np.fromfile(fh, count=1, dtype=">f4") # NOQA
hubble = np.fromfile(fh, count=1, dtype=">f4") # NOQA
Wp5 = np.fromfile(fh, count=1, dtype=">f4") # NOQA
Ocurv = np.fromfile(fh, count=1, dtype=">f4") # NOQA
wspecies = np.fromfile(fh, count=10, dtype=">f4") # NOQA
lspecies = np.fromfile(fh, count=10, dtype=">i4") # NOQA
extras = np.fromfile(fh, count=79, dtype=">f4") # NOQA
boxsize = np.fromfile(fh, count=1, dtype=">f4") # NOQA
return True
except Exception:
return False
return False
ref_sqrt = numpy.zeros((128, 128))
bg_seq = numpy.zeros((128, 128, positions))
#call .exe file to do imaging
#subprocess.call([r"qcl-holo-cap.exe", data_dir, '--zstep', str(zstep), '--positions', str(positions), '--frames', str(num_frames), '--laserpower', str(lp), '--inte', str(inte)])
#read in reference image
re_dir = r'C:\Users\shihao\Desktop\ir-images\ir-holography\bead5\ref'
re_list = os.listdir(re_dir)
number_res = len(re_list)
if( number_res != 20):
print("Error: Reference Image is missing!")
else:
for i in range(number_res):
ref_dir = re_dir + '\\' + re_list[i]
ref_data = numpy.fromfile(ref_dir, dtype = numpy.uint16, count = -1, sep = '')
ref_data = numpy.reshape(ref_data, (128, 128), order = 'C')
ref_seq[:, :, i] = ref_data
ref_intensity = sum(ref_seq, axis = 2) / positions
ref_sqrt = numpy.sqrt(ref_intensity)
#read in background image
bgs_dir = r'C:\Users\shihao\Desktop\ir-images\ir-holography\bead5\bg_holo'
bgs_list = os.listdir(bgs_dir)
number_bgs = len(bgs_list)
if( number_bgs != positions):
print("Error: Background Image is missing!")
else:
for i in range(number_bgs):
bg_dir = bgs_dir + '\\' + bgs_list[i]
def _parse_parameter_file(self):
"""
Get the various simulation parameters & constants.
"""
self.domain_left_edge = np.zeros(3, dtype="float")
self.domain_right_edge = np.zeros(3, dtype="float") + 1.0
self.dimensionality = 3
self.refine_by = 2
self._periodicity = (True, True, True)
self.cosmological_simulation = True
self.parameters = {}
self.parameters.update(constants)
self.parameters["Time"] = 1.0
# read the amr header
with open(self._file_amr, "rb") as f:
amr_header_vals = fpu.read_attrs(f, amr_header_struct, ">")
n_to_skip = len(("tl", "dtl", "tlold", "dtlold", "iSO"))
fpu.skip(f, n_to_skip, endian=">")
(self.ncell) = fpu.read_vector(f, "i", ">")[0]
# Try to figure out the root grid dimensions
est = int(np.rint(self.ncell ** (1.0 / 3.0)))
# Note here: this is the number of *cells* on the root grid.
# This is not the same as the number of Octs.
# domain dimensions is the number of root *cells*
self.domain_dimensions = np.ones(3, dtype="int64") * est
self.root_grid_mask_offset = f.tell()
self.root_nocts = self.domain_dimensions.prod() // 8
self.root_ncells = self.root_nocts * 8
mylog.debug(
"Estimating %i cells on a root grid side, %i root octs",
est,
self.root_nocts,
)
self.root_iOctCh = fpu.read_vector(f, "i", ">")[: self.root_ncells]
self.root_iOctCh = self.root_iOctCh.reshape(
self.domain_dimensions, order="F"
)
self.root_grid_offset = f.tell()
self.root_nhvar = fpu.skip(f, endian=">")
self.root_nvar = fpu.skip(f, endian=">")
# make sure that the number of root variables is a multiple of
# rootcells
assert self.root_nhvar % self.root_ncells == 0
assert self.root_nvar % self.root_ncells == 0
self.nhydro_variables = (
self.root_nhvar + self.root_nvar
) / self.root_ncells
self.iOctFree, self.nOct = fpu.read_vector(f, "i", ">")
self.child_grid_offset = f.tell()
# lextra needs to be loaded as a string, but it's actually
# array values. So pop it off here, and then re-insert.
lextra = amr_header_vals.pop("lextra")
amr_header_vals["lextra"] = np.frombuffer(lextra, ">f4")
self.parameters.update(amr_header_vals)
amr_header_vals = None
# estimate the root level
float_center, fl, iocts, nocts, root_level = _read_art_level_info(
f, [0, self.child_grid_offset], 1, coarse_grid=self.domain_dimensions[0]
)
del float_center, fl, iocts, nocts
self.root_level = root_level
mylog.info("Using root level of %02i", self.root_level)
# read the particle header
self.particle_types = []
self.particle_types_raw = ()
if not self.skip_particles and self._file_particle_header:
with open(self._file_particle_header, "rb") as fh:
particle_header_vals = fpu.read_attrs(fh, particle_header_struct, ">")
fh.seek(seek_extras)
n = particle_header_vals["Nspecies"]
wspecies = np.fromfile(fh, dtype=">f", count=10)
lspecies = np.fromfile(fh, dtype=">i", count=10)
# extras needs to be loaded as a string, but it's actually
# array values. So pop it off here, and then re-insert.
extras = particle_header_vals.pop("extras")
particle_header_vals["extras"] = np.frombuffer(extras, ">f4")
self.parameters["wspecies"] = wspecies[:n]
self.parameters["lspecies"] = lspecies[:n]
for specie in range(n):
self.particle_types.append("specie%i" % specie)
self.particle_types_raw = tuple(self.particle_types)
ls_nonzero = np.diff(lspecies)[: n - 1]
ls_nonzero = np.append(lspecies[0], ls_nonzero)
self.star_type = len(ls_nonzero)
mylog.info("Discovered %i species of particles", len(ls_nonzero))
info_str = "Particle populations: " + "%9i " * len(ls_nonzero)
mylog.info(info_str, *ls_nonzero)
self._particle_type_counts = dict(zip(self.particle_types_raw, ls_nonzero))
for k, v in particle_header_vals.items():
if k in self.parameters.keys():
if not self.parameters[k] == v:
mylog.info(
"Inconsistent parameter %s %1.1e %1.1e",
k,
v,
self.parameters[k],
)
else:
self.parameters[k] = v
self.parameters_particles = particle_header_vals
self.parameters.update(particle_header_vals)
self.parameters["ng"] = self.parameters["Ngridc"]
self.parameters["ncell0"] = self.parameters["ng"] ** 3
# setup standard simulation params yt expects to see
self.current_redshift = self.parameters["aexpn"] ** -1.0 - 1.0
self.omega_lambda = self.parameters["Oml0"]
self.omega_matter = self.parameters["Om0"]
self.hubble_constant = self.parameters["hubble"]
self.min_level = self.parameters["min_level"]
self.max_level = self.parameters["max_level"]
if self.limit_level is not None:
self.max_level = min(self.limit_level, self.parameters["max_level"])
if self.force_max_level is not None:
self.max_level = self.force_max_level
self.hubble_time = 1.0 / (self.hubble_constant * 100 / 3.08568025e19)
self.current_time = self.quan(b2t(self.parameters["t"]), "Gyr")
self.gamma = self.parameters["gamma"]
mylog.info("Max level is %02i", self.max_level)
def restore_weights(self, weights_file):
"""Load previously trained model weights
Arguments:
weights_file: beginning by project root this is the path
where is save your weights; example: "weights/weights_01.h5"
"""
tf.keras.backend.clear_session() # used to reset layer names
# load Darknet original weights to TensorFlow model
range1 = 75 if self.version == "yolov3" else 13
range2 = [58, 66, 74] if self.version == "yolov3" else [9, 12]
with open(weights_file, 'rb') as wf:
major, minor, revision, seen, _ = np.fromfile(wf,
dtype=np.int32,
count=5)
j = 0
for i in range(range1):
if i > 0:
conv_layer_name = 'conv2d_%d' % i
else:
conv_layer_name = 'conv2d'
if j > 0:
bn_layer_name = 'batch_normalization_%d' % j
else:
bn_layer_name = 'batch_normalization'
conv_layer = self.model.get_layer(conv_layer_name)
filters = conv_layer.filters
k_size = conv_layer.kernel_size[0]
in_dim = conv_layer.input_shape[-1]
if i not in range2:
# darknet weights: [beta, gamma, mean, variance]
bn_weights = np.fromfile(wf,
dtype=np.float32,
count=4 * filters)
# tf weights: [gamma, beta, mean, variance]
bn_weights = bn_weights.reshape((4, filters))[[1, 0, 2, 3]]
bn_layer = self.model.get_layer(bn_layer_name)
j += 1
else:
conv_bias = np.fromfile(wf,
dtype=np.float32,
count=filters)
# darknet shape (out_dim, in_dim, height, width)
conv_shape = (filters, in_dim, k_size, k_size)
conv_weights = np.fromfile(wf,
dtype=np.float32,
count=np.product(conv_shape))
# tf shape (height, width, in_dim, out_dim)
conv_weights = conv_weights.reshape(conv_shape).transpose(
[2, 3, 1, 0])
if i not in range2:
conv_layer.set_weights([conv_weights])
bn_layer.set_weights(bn_weights)
else:
conv_layer.set_weights([conv_weights, conv_bias])
assert len(wf.read()) == 0, 'failed to read all data'
def readTrc( fName ):
"""
Reads .trc binary files from LeCroy Oscilloscopes.
Decoding is based on LECROY_2_3 template.
[More info](http://forums.ni.com/attachments/ni/60/4652/2/LeCroyWaveformTemplate_2_3.pdf)
Parameters
-----------
fName = filename of the .trc file
Returns
-----------
x: array with sample times [s],
y: array with sample values [V],
d: dictionary with metadata
M. Betz 09/2015
"""
with open(fName, "rb") as fid:
data = fid.read(50).decode()
wdOffset = data.find('WAVEDESC')
#------------------------
# Get binary format / endianess
#------------------------
if readX( fid, '?', wdOffset + 32 ): #16 or 8 bit sample format?
smplFmt = "int16"
else:
smplFmt = "int8"
if readX( fid, '?', wdOffset + 34 ): #Big or little endian?
endi = "<"
else:
endi = ">"
#------------------------
# Get length of blocks and arrays:
#------------------------
lWAVE_DESCRIPTOR = readX( fid, endi+"l", wdOffset + 36 )
lUSER_TEXT = readX( fid, endi+"l", wdOffset + 40 )
lTRIGTIME_ARRAY = readX( fid, endi+"l", wdOffset + 48 )
lRIS_TIME_ARRAY = readX( fid, endi+"l", wdOffset + 52 )
lWAVE_ARRAY_1 = readX( fid, endi+"l", wdOffset + 60 )
lWAVE_ARRAY_2 = readX( fid, endi+"l", wdOffset + 64 )
d = dict() #Will store all the extracted Metadata
#------------------------
# Get Instrument info
#------------------------
d["INSTRUMENT_NAME"] = readX( fid, "16s", wdOffset + 76 ).decode().split('\x00')[0]
d["INSTRUMENT_NUMBER"]= readX( fid, endi+"l", wdOffset + 92 )
d["TRACE_LABEL"] = readX( fid, "16s", wdOffset + 96 ).decode().split('\x00')[0]
#------------------------
# Get Waveform info
#------------------------
d["WAVE_ARRAY_COUNT"] = readX( fid, endi+"l", wdOffset +116 )
d["PNTS_PER_SCREEN"] = readX( fid, endi+"l", wdOffset +120 )
d["FIRST_VALID_PNT"] = readX( fid, endi+"l", wdOffset +124 )
d["LAST_VALID_PNT"] = readX( fid, endi+"l", wdOffset +128 )
d["FIRST_POINT"] = readX( fid, endi+"l", wdOffset +132 )
d["SPARSING_FACTOR"] = readX( fid, endi+"l", wdOffset +136 )
d["SEGMENT_INDEX"] = readX( fid, endi+"l", wdOffset +140 )
d["SUBARRAY_COUNT"] = readX( fid, endi+"l", wdOffset +144 )
d["SWEEPS_PER_ACQ"] = readX( fid, endi+"l", wdOffset +148 )
d["POINTS_PER_PAIR"] = readX( fid, endi+"h", wdOffset +152 )
d["PAIR_OFFSET"] = readX( fid, endi+"h", wdOffset +154 )
d["VERTICAL_GAIN"] = readX( fid, endi+"f", wdOffset +156 ) #to get floating values from raw data :
d["VERTICAL_OFFSET"] = readX( fid, endi+"f", wdOffset +160 ) #VERTICAL_GAIN * data - VERTICAL_OFFSET
d["MAX_VALUE"] = readX( fid, endi+"f", wdOffset +164 )
d["MIN_VALUE"] = readX( fid, endi+"f", wdOffset +168 )
d["NOMINAL_BITS"] = readX( fid, endi+"h", wdOffset +172 )
d["NOM_SUBARRAY_COUNT"]= readX( fid, endi+"h",wdOffset +174 )
d["HORIZ_INTERVAL"] = readX( fid, endi+"f", wdOffset +176 ) #sampling interval for time domain waveforms
d["HORIZ_OFFSET"] = readX( fid, endi+"d", wdOffset +180 ) #trigger offset for the first sweep of the trigger, seconds between the trigger and the first data point
d["PIXEL_OFFSET"] = readX( fid, endi+"d", wdOffset +188 )
d["VERTUNIT"] = readX( fid, "48s", wdOffset +196 ).decode().split('\x00')[0]
d["HORUNIT"] = readX( fid, "48s", wdOffset +244 ).decode().split('\x00')[0]
d["HORIZ_UNCERTAINTY"]= readX( fid, endi+"f", wdOffset +292 )
d["TRIGGER_TIME"] = getTimeStamp( fid, endi, wdOffset +296 )
d["ACQ_DURATION"] = readX( fid, endi+"f", wdOffset +312 )
d["RECORD_TYPE"] = ["single_sweep","interleaved","histogram","graph","filter_coefficient","complex","extrema","sequence_obsolete","centered_RIS","peak_detect"][ readX( fid, endi+"H", wdOffset +316 ) ]
d["PROCESSING_DONE"] = ["no_processing","fir_filter","interpolated","sparsed","autoscaled","no_result","rolling","cumulative"][ readX( fid, endi+"H", wdOffset +318 ) ]
d["RIS_SWEEPS"] = readX( fid, endi+"h", wdOffset +322 )
d["TIMEBASE"] = ['1_ps/div', '2_ps/div', '5_ps/div', '10_ps/div', '20_ps/div', '50_ps/div', '100_ps/div', '200_ps/div', '500_ps/div', '1_ns/div', '2_ns/div', '5_ns/div', '10_ns/div', '20_ns/div', '50_ns/div', '100_ns/div', '200_ns/div', '500_ns/div', '1_us/div', '2_us/div', '5_us/div', '10_us/div', '20_us/div', '50_us/div', '100_us/div', '200_us/div', '500_us/div', '1_ms/div', '2_ms/div', '5_ms/div', '10_ms/div', '20_ms/div', '50_ms/div', '100_ms/div', '200_ms/div', '500_ms/div', '1_s/div', '2_s/div', '5_s/div', '10_s/div', '20_s/div', '50_s/div', '100_s/div', '200_s/div', '500_s/div', '1_ks/div', '2_ks/div', '5_ks/div', 'EXTERNAL'][ readX( fid, endi+"H", wdOffset +324 ) ]
d["VERT_COUPLING"] = ['DC_50_Ohms', 'ground', 'DC_1MOhm', 'ground', 'AC,_1MOhm'][ readX( fid, endi+"H", wdOffset +326 ) ]
d["PROBE_ATT"] = readX( fid, endi+"f", wdOffset +328 )
d["FIXED_VERT_GAIN"] = ['1_uV/div','2_uV/div','5_uV/div','10_uV/div','20_uV/div','50_uV/div','100_uV/div','200_uV/div','500_uV/div','1_mV/div','2_mV/div','5_mV/div','10_mV/div','20_mV/div','50_mV/div','100_mV/div','200_mV/div','500_mV/div','1_V/div','2_V/div','5_V/div','10_V/div','20_V/div','50_V/div','100_V/div','200_V/div','500_V/div','1_kV/div'][ readX( fid, endi+"H", wdOffset +332 ) ]
d["BANDWIDTH_LIMIT"] = ['off', 'on'][ readX( fid, endi+"H", wdOffset +334 ) ]
d["VERTICAL_VERNIER"] = readX( fid, endi+"f", wdOffset +336 )
d["ACQ_VERT_OFFSET"] = readX( fid, endi+"f", wdOffset +340 )
d["WAVE_SOURCE"] = readX( fid, endi+"H", wdOffset +344 )
d["USER_TEXT"] = readX( fid, "{0}s".format(lUSER_TEXT), wdOffset + lWAVE_DESCRIPTOR ).decode().split('\x00')[0]
#------------------------
# Get main sample data with the help of numpys .fromfile(
#------------------------
fid.seek( wdOffset + lWAVE_DESCRIPTOR + lUSER_TEXT + lTRIGTIME_ARRAY + lRIS_TIME_ARRAY ) #Seek to WAVE_ARRAY_1
y = np.fromfile( fid, smplFmt, lWAVE_ARRAY_1 )
if endi == ">":
y.byteswap( True )
y = d["VERTICAL_GAIN"] * y - d["VERTICAL_OFFSET"]
x = np.arange(1,len(y)+1)*d["HORIZ_INTERVAL"] + d["HORIZ_OFFSET"]
return x, y, d
def _parse_parameter_file(self):
"""
Get the various simulation parameters & constants.
"""
self.domain_left_edge = np.zeros(3, dtype="float")
self.domain_right_edge = np.zeros(3, dtype="float") + 1.0
self.dimensionality = 3
self.refine_by = 2
self._periodicity = (True, True, True)
self.cosmological_simulation = True
self.parameters = {}
self.parameters.update(constants)
self.parameters["Time"] = 1.0
self.file_count = 1
self.filename_template = self.parameter_filename
# read the particle header
self.particle_types = []
self.particle_types_raw = ()
assert self._file_particle_header
with open(self._file_particle_header, "rb") as fh:
seek = 4
fh.seek(seek)
headerstr = fh.read(45).decode("ascii")
aexpn = np.fromfile(fh, count=1, dtype=">f4")
aexp0 = np.fromfile(fh, count=1, dtype=">f4")
amplt = np.fromfile(fh, count=1, dtype=">f4")
astep = np.fromfile(fh, count=1, dtype=">f4")
istep = np.fromfile(fh, count=1, dtype=">i4")
partw = np.fromfile(fh, count=1, dtype=">f4")
tintg = np.fromfile(fh, count=1, dtype=">f4")
ekin = np.fromfile(fh, count=1, dtype=">f4")
ekin1 = np.fromfile(fh, count=1, dtype=">f4")
ekin2 = np.fromfile(fh, count=1, dtype=">f4")
au0 = np.fromfile(fh, count=1, dtype=">f4")
aeu0 = np.fromfile(fh, count=1, dtype=">f4")
nrowc = np.fromfile(fh, count=1, dtype=">i4")
ngridc = np.fromfile(fh, count=1, dtype=">i4")
nspecs = np.fromfile(fh, count=1, dtype=">i4")
nseed = np.fromfile(fh, count=1, dtype=">i4")
Om0 = np.fromfile(fh, count=1, dtype=">f4")
Oml0 = np.fromfile(fh, count=1, dtype=">f4")
hubble = np.fromfile(fh, count=1, dtype=">f4")
Wp5 = np.fromfile(fh, count=1, dtype=">f4")
Ocurv = np.fromfile(fh, count=1, dtype=">f4")
wspecies = np.fromfile(fh, count=10, dtype=">f4")
lspecies = np.fromfile(fh, count=10, dtype=">i4")
extras = np.fromfile(fh, count=79, dtype=">f4")
boxsize = np.fromfile(fh, count=1, dtype=">f4")
n = nspecs[0]
particle_header_vals = {}
tmp = [
headerstr,
aexpn,
aexp0,
amplt,
astep,
istep,
partw,
tintg,
ekin,
ekin1,
ekin2,
au0,
aeu0,
nrowc,
ngridc,
nspecs,
nseed,
Om0,
Oml0,
hubble,
Wp5,
Ocurv,
wspecies,
lspecies,
extras,
boxsize,
]
for i, arr in enumerate(tmp):
a1 = dmparticle_header_struct[0][i]
a2 = dmparticle_header_struct[1][i]
if a2 == 1:
particle_header_vals[a1] = arr[0]
else:
particle_header_vals[a1] = arr[:a2]
for specie in range(n):
self.particle_types.append("specie%i" % specie)
self.particle_types_raw = tuple(self.particle_types)
ls_nonzero = np.diff(lspecies)[: n - 1]
ls_nonzero = np.append(lspecies[0], ls_nonzero)
self.star_type = len(ls_nonzero)
mylog.info("Discovered %i species of particles", len(ls_nonzero))
info_str = "Particle populations: " + "%9i " * len(ls_nonzero)
mylog.info(info_str, *ls_nonzero)
for k, v in particle_header_vals.items():
if k in self.parameters.keys():
if not self.parameters[k] == v:
mylog.info(
"Inconsistent parameter %s %1.1e %1.1e",
k,
v,
self.parameters[k],
)
else:
self.parameters[k] = v
self.parameters_particles = particle_header_vals
self.parameters.update(particle_header_vals)
self.parameters["wspecies"] = wspecies[:n]
self.parameters["lspecies"] = lspecies[:n]
self.parameters["ng"] = self.parameters["Ngridc"]
self.parameters["ncell0"] = self.parameters["ng"] ** 3
self.parameters["boxh"] = self.parameters["boxsize"]
self.parameters["total_particles"] = ls_nonzero
self.domain_dimensions = np.ones(3, dtype="int64") * 2 # NOT ng
# setup standard simulation params yt expects to see
# Convert to float to please unyt
self.current_redshift = float(self.parameters["aexpn"] ** -1.0 - 1.0)
self.omega_lambda = float(particle_header_vals["Oml0"])
self.omega_matter = float(particle_header_vals["Om0"])
self.hubble_constant = float(particle_header_vals["hubble"])
self.min_level = 0
self.max_level = 0
# self.min_level = particle_header_vals['min_level']
# self.max_level = particle_header_vals['max_level']
# if self.limit_level is not None:
# self.max_level = min(
# self.limit_level, particle_header_vals['max_level'])
# if self.force_max_level is not None:
# self.max_level = self.force_max_level
self.hubble_time = 1.0 / (self.hubble_constant * 100 / 3.08568025e19)
self.parameters["t"] = a2b(self.parameters["aexpn"])
self.current_time = self.quan(b2t(self.parameters["t"]), "Gyr")
self.gamma = self.parameters["gamma"]
mylog.info("Max level is %02i", self.max_level)
def gen_data_dx(fmap_shape,
filter_shape,
pad_,
stride_,
dilation_,
expect_file,
attrs=None):
block_size = 16
in_n, in_c, in_h, in_w = fmap_shape
cout, cin, w_h, w_w = filter_shape
assert in_c == cin
in_c = (in_c + block_size - 1) // block_size * block_size
cout = (cout + block_size - 1) // block_size * block_size
pad_top, pad_bottom, pad_left, pad_right = pad_
stride_h, stride_w = stride_
dilation_h, dilation_w = dilation_
assert dilation_h == 1
assert dilation_w == 1
x_shape = (in_n, in_c, in_h, in_w)
w_shape = (cout, in_c, w_h, w_w)
b_shape = (w_shape[0], )
p_top = w_h - pad_top - 1
p_left = w_w - pad_left - 1
p_bottom = in_h + pad_top - stride_h * (
(in_h + pad_top + pad_bottom - w_h) // stride_h + 1)
p_right = in_w + pad_left - stride_w * (
(in_w + pad_left + pad_right - w_w) // stride_w + 1)
print("Data gen ...")
x = random_gaussian(x_shape, miu=1, sigma=0.1).astype(np.float16)
w = random_gaussian(w_shape, miu=1, sigma=0.1).astype(np.float16)
Ho = (x_shape[2] + pad_top + pad_bottom - w_shape[2]) // stride_h + 1
Wo = (x_shape[3] + pad_left + pad_right - w_shape[3]) // stride_w + 1
out_shape = (x_shape[0], w_shape[0], Ho, Wo)
dout = random_gaussian(out_shape, miu=1, sigma=0.1).astype(np.float16)
dx_shape = (in_n, in_c // block_size, in_h, in_w, block_size)
flag_w = os.environ.get("WRITE_TO_DISK", "No")
if (flag_w == "No") and (os.path.exists(expect_file) == True):
# read expect from file
dx = np.fromfile(expect_file, np.float16).reshape(dx_shape)
else:
# compute expect data:
dx = calculate_conv_backprop_input(x, w, dout,
[p_top, p_bottom, p_left, p_right],
[stride_h, stride_w])
if flag_w == "Yes":
# write expect to file
with open(expect_file, "w+") as file:
dx.tofile(file)
file.close()
# reshape
C0 = block_size
ON, OC, OH, OW = out_shape
WN, WC, WH, WW = w_shape
dout = dout.reshape(ON, OC // C0, C0, OH, OW).transpose(0, 1, 3, 4,
2).copy()
w = w.reshape(WN, WC // C0, C0, WH, WW).transpose(1, 3, 4, 0, 2).copy()
return dout, w, dx
import numpy as np
import matplotlib.pyplot as plt
import KleinFunction as Klein
image_size = 28
no_of_different_labels = 10
image_pixels = image_size*image_size
usebinQ = True
if(usebinQ is False):
#train_data = np.loadtxt("../../mnist_train.csv",delimiter=",")
train_data = np.loadtxt("../../mnist_train_100.csv",delimiter=",")
fac = 255 # normalising data values to [0., 1.]
train_imgs = np.asfarray(train_data[:, 1:]) / fac
train_labels = np.asfarray(train_data[:, :1])
else:
train_imgs = np.fromfile("../../mnist_train_imgs_binary.dat").reshape((60000,image_pixels))
train_labels = np.fromfile("../../mnist_train_labels_binary.dat").reshape((60000,1))
test_imgs = np.fromfile("../../mnist_test_imgs_binary.dat").reshape((10000,image_pixels))
test_labels = np.fromfile("../../mnist_test_labels_binary.dat").reshape((10000,1))
no_of_hidden_nodes = 100
accuracy = 0.96
epochs = 100#30
bias = None
nAlgorithm = 1 # excluding minibatch
run_minibatch = False
arr_learning_rate = [0.001,0.002,0.005,0.01,0.02,0.05,0.1,0.2,0.5,1.0]
save_frequency = 6000
threshold = 0.0005
arr_momentum_para = [0.,0.9,0.] # 0.9 is for the algorithm with momentum
ebs = [i * 1e-3 for i in range(1, 10, 2)]
else:
ebs = [i * 1e-2 for i in range(1, 8, 2)] + [0.1]
idxrange = [0]
pid = str(os.getpid()).strip()
data = np.zeros((len(ebs) + 1, len(idxrange) + 1, 9), dtype=np.float32)
for i in range(9):
data[1:, 0, i] = ebs
data[0, 1:, i] = idxrange
for i, idx in enumerate(idxrange):
filename = "%s.dat.log10" % field
filepath = os.path.join(datafolder, filename)
a = np.fromfile(filepath, dtype=np.float32)
a = a - np.min(a)
filepath = filepath + ".positive"
a.tofile(filepath)
for j, eb in enumerate(ebs):
os.system("python3 Autoencoder_Prototype.py -c %s -e %f -n %s" %
(filepath, eb, field))
zpath = filepath + ".z"
dpath = zpath + ".d"
dvname = filepath.split("/")[-1] + ".dvalue"
dvpath = filepath + ".dvalue"
os.system("du -s %s*&>%s.txt" % (filepath, pid))
origsize = 0
compressedsize = 0
with open("%s.txt" % pid, "r") as f:
lines = f.read().splitlines()
def load_weights(self, weightfile):
#Open the weights file
fp = open(weightfile, "rb")
#The first 5 values are header information
# 1. Major version number
# 2. Minor Version Number
# 3. Subversion number
# 4,5. Images seen by the network (during training)
header = np.fromfile(fp, dtype=np.int32, count=5)
self.header = torch.from_numpy(header)
self.seen = self.header[3]
weights = np.fromfile(fp, dtype=np.float32)
ptr = 0
for i in range(len(self.module_list)):
module_type = self.blocks[i + 1]["type"]
#If module_type is convolutional load weights
#Otherwise ignore.
if module_type == "convolutional":
model = self.module_list[i]
try:
batch_normalize = int(self.blocks[i +
1]["batch_normalize"])
except:
batch_normalize = 0
conv = model[0]
if (batch_normalize):
bn = model[1]
#Get the number of weights of Batch Norm Layer
num_bn_biases = bn.bias.numel()
#Load the weights
bn_biases = torch.from_numpy(weights[ptr:ptr +
num_bn_biases])
ptr += num_bn_biases
bn_weights = torch.from_numpy(weights[ptr:ptr +
num_bn_biases])
ptr += num_bn_biases
bn_running_mean = torch.from_numpy(weights[ptr:ptr +
num_bn_biases])
ptr += num_bn_biases
bn_running_var = torch.from_numpy(weights[ptr:ptr +
num_bn_biases])
ptr += num_bn_biases
#Cast the loaded weights into dims of model weights.
bn_biases = bn_biases.view_as(bn.bias.data)
bn_weights = bn_weights.view_as(bn.weight.data)
bn_running_mean = bn_running_mean.view_as(bn.running_mean)
bn_running_var = bn_running_var.view_as(bn.running_var)
#Copy the data to model
bn.bias.data.copy_(bn_biases)
bn.weight.data.copy_(bn_weights)
bn.running_mean.copy_(bn_running_mean)
bn.running_var.copy_(bn_running_var)
else:
#Number of biases
num_biases = conv.bias.numel()
#Load the weights
conv_biases = torch.from_numpy(weights[ptr:ptr +
num_biases])
ptr = ptr + num_biases
#reshape the loaded weights according to the dims of the model weights
conv_biases = conv_biases.view_as(conv.bias.data)
#Finally copy the data
conv.bias.data.copy_(conv_biases)
#Let us load the weights for the Convolutional layers
num_weights = conv.weight.numel()
#Do the same as above for weights
conv_weights = torch.from_numpy(weights[ptr:ptr + num_weights])
ptr = ptr + num_weights
conv_weights = conv_weights.view_as(conv.weight.data)
conv.weight.data.copy_(conv_weights)
def read_all_images(path_to_data):
with open(path_to_data, 'rb') as f:
everything = np.fromfile(f, dtype=np.uint8)
images = np.reshape(everything, (-1, 3, 96, 96))
images = np.transpose(images, (0, 3, 2, 1))
return images
def load_weights(var_list, weights_file):
"""
Loads and converts pre-trained weights.
param:
var_list: list of network variables.
weights_file: name of the binary file.
"""
with open(weights_file, "rb") as fp:
np.fromfile(fp, dtype=np.int32, count=5)
weights = np.fromfile(fp, dtype=np.float32)
ptr = 0
i = 0
assign_ops = []
while i < len(var_list) - 1:
try:
var1 = var_list[i]
var2 = var_list[i + 1]
# do something only if we process conv layer
if 'Conv' in var1.name.split('/')[-2]:
# check type of next layer
if 'BatchNorm' in var2.name.split('/')[-2]:
# load batch norm params
gamma, beta, mean, var = var_list[i + 1:i + 5]
batch_norm_vars = [beta, gamma, mean, var]
for var in batch_norm_vars:
shape = var.shape.as_list()
num_params = np.prod(shape)
var_weights = weights[ptr:ptr + num_params].reshape(shape)
ptr += num_params
assign_ops.append(tf.assign(var, var_weights, validate_shape=True))
# we move the pointer by 4, because we loaded 4 variables
i += 4
elif 'Conv' in var2.name.split('/')[-2]:
# load biases
bias = var2
bias_shape = bias.shape.as_list()
bias_params = np.prod(bias_shape)
bias_weights = weights[ptr:ptr +
bias_params].reshape(bias_shape)
ptr += bias_params
assign_ops.append(tf.assign(bias, bias_weights, validate_shape=True))
# we loaded 1 variable
i += 1
# we can load weights of conv layer
shape = var1.shape.as_list()
num_params = np.prod(shape)
var_weights = weights[ptr:ptr + num_params].reshape(
(shape[3], shape[2], shape[0], shape[1]))
# remember to transpose to column-major
var_weights = np.transpose(var_weights, (2, 3, 1, 0))
ptr += num_params
assign_ops.append(
tf.assign(var1, var_weights, validate_shape=True))
i += 1
except ValueError:
print("tensor", i)
return assign_ops
def read_mrc(path, read_data=True, show_progress=False):
path = os.path.realpath(path)
with open(path, 'rb') as f:
mrc = {}
mrc['nx'] = int(struct.unpack('i', f.read(4))[0])
mrc['ny'] = int(struct.unpack('i', f.read(4))[0])
mrc['nz'] = int(struct.unpack('i', f.read(4))[0])
mrc['mode'] = struct.unpack('i', f.read(4))[0]
mrc['nxstart'] = struct.unpack('i', f.read(4))[0]
mrc['nystart'] = struct.unpack('i', f.read(4))[0]
mrc['nzstart'] = struct.unpack('i', f.read(4))[0]
mrc['mx'] = struct.unpack('i', f.read(4))[0]
mrc['my'] = struct.unpack('i', f.read(4))[0]
mrc['mz'] = struct.unpack('i', f.read(4))[0]
mrc['xlen'] = struct.unpack('f', f.read(4))[0]
mrc['ylen'] = struct.unpack('f', f.read(4))[0]
mrc['zlen'] = struct.unpack('f', f.read(4))[0]
mrc['alpha'] = struct.unpack('f', f.read(4))[0]
mrc['beta'] = struct.unpack('f', f.read(4))[0]
mrc['gamma'] = struct.unpack('f', f.read(4))[0]
mrc['mapc'] = struct.unpack('i', f.read(4))[0]
mrc['mapr'] = struct.unpack('i', f.read(4))[0]
mrc['maps'] = struct.unpack('i', f.read(4))[0]
mrc['amin'] = struct.unpack('f', f.read(4))[0]
mrc['amax'] = struct.unpack('f', f.read(4))[0]
mrc['amean'] = struct.unpack('f', f.read(4))[0]
mrc['ispg'] = struct.unpack('h', f.read(2))[0]
mrc['nsymbt'] = struct.unpack('h', f.read(2))[0]
mrc['next'] = struct.unpack('i', f.read(4))[0]
mrc['creatid'] = struct.unpack('h', f.read(2))[0]
mrc['unused1'] = struct.unpack(('c' * 30), f.read(30))[0]
mrc['nint'] = struct.unpack('h', f.read(2))[0]
mrc['nreal'] = struct.unpack('h', f.read(2))[0]
mrc['unused2'] = struct.unpack(('c' * 28), f.read(28))[0]
mrc['idtype'] = struct.unpack('h', f.read(2))[0]
mrc['lens'] = struct.unpack('h', f.read(2))[0]
mrc['nd1'] = struct.unpack('h', f.read(2))[0]
mrc['nd2'] = struct.unpack('h', f.read(2))[0]
mrc['vd1'] = struct.unpack('h', f.read(2))[0]
mrc['vd2'] = struct.unpack('h', f.read(2))[0]
mrc['tiltangles'] = struct.unpack(('f' * 6), f.read((4 * 6)))
mrc['xorg'] = struct.unpack('f', f.read(4))[0]
mrc['yorg'] = struct.unpack('f', f.read(4))[0]
mrc['zorg'] = struct.unpack('f', f.read(4))[0]
mrc['cmap'] = struct.unpack(('c' * 4), f.read(4))
mrc['stamp'] = struct.unpack(('c' * 4), f.read(4))
mrc['rms'] = struct.unpack('f', f.read(4))[0]
mrc['nlabl'] = struct.unpack('i', f.read(4))[0]
mrc['labl'] = struct.unpack(('c' * 800), f.read(800))
size = [mrc['nx'], mrc['ny'], mrc['nz']]
n_voxel = N.prod(size)
extended = {}
extended['magnification'] = [0]
extended['exp_time'] = [0]
extended['pixelsize'] = [0]
extended['defocus'] = [0]
extended['a_tilt'] = ([0] * mrc['nz'])
extended['tiltaxis'] = [0]
if (mrc['next'] != 0):
nbh = (mrc['next'] / 128)
if (nbh == 1024):
for lauf in range(nbh):
extended['a_tilt'][lauf] = struct.unpack('f', f.read(4))[0]
extended['b_tilt'][lauf] = struct.unpack('f', f.read(4))[0]
extended['x_stage'][lauf] = struct.unpack('f',
f.read(4))[0]
extended['y_stage'][lauf] = struct.unpack('f',
f.read(4))[0]
extended['z_stage'][lauf] = struct.unpack('f',
f.read(4))[0]
extended['x_shift'][lauf] = struct.unpack('f',
f.read(4))[0]
extended['y_shift'][lauf] = struct.unpack('f',
f.read(4))[0]
extended['defocus'][lauf] = struct.unpack('f',
f.read(4))[0]
extended['exp_time'][lauf] = struct.unpack('f',
f.read(4))[0]
extended['mean_int'][lauf] = struct.unpack('f',
f.read(4))[0]
extended['tiltaxis'][lauf] = struct.unpack('f',
f.read(4))[0]
extended['tiltaxis'][lauf] = struct.unpack('f',
f.read(4))[0]
extended['pixelsize'][lauf] = struct.unpack(
'f', f.read(4))[0]
extended['magnification'][lauf] = struct.unpack(
'f', f.read(4))[0]
f.seek(offset=(128 - 52), whence=1)
else:
f.seek(offset=MRC.__next__, whence=1)
if read_data:
slice_voxel_num = (mrc['nx'] * mrc['ny'])
v = None
for i in range(mrc['nz']):
if show_progress:
print('\r', i, ' ', end=' ')
sys.stdout.flush()
if (mrc['mode'] == 0):
if (v is None):
v = N.zeros(size, dtype=N.int8)
data_read = N.fromfile(f,
dtype=N.int8,
count=slice_voxel_num)
elif (mrc['mode'] == 1):
if (v is None):
v = N.zeros(size, dtype=N.int16)
data_read = N.fromfile(f,
dtype=N.int16,
count=slice_voxel_num)
elif (mrc['mode'] == 2):
if (v is None):
v = N.zeros(size, dtype=N.float32)
data_read = N.fromfile(f,
dtype=N.float32,
count=slice_voxel_num)
else:
raise Exception(
'Sorry, i cannot read this as an MRC-File !!!')
data_read = None
if (data_read.size != slice_voxel_num):
import pdb
pdb.set_trace()
v[:, :, i] = N.reshape(data_read, (mrc['nx'], mrc['ny']),
order='F')
else:
v = None
h = {}
h['Voltage'] = None
h['Cs'] = None
h['Aperture'] = None
h['Magnification'] = extended['magnification'][0]
h['Postmagnification'] = None
h['Exposuretime'] = extended['exp_time'][0]
h['Objectpixelsize'] = (extended['pixelsize'][0] * 1000000000.0)
h['Microscope'] = None
h['Pixelsize'] = None
h['CCDArea'] = None
h['Defocus'] = extended['defocus'][0]
h['Astigmatism'] = None
h['AstigmatismAngle'] = None
h['FocusIncrement'] = None
h['CountsPerElectron'] = None
h['Intensity'] = None
h['EnergySlitwidth'] = None
h['EnergyOffset'] = None
h['Tiltangle'] = extended['a_tilt'][:mrc['nz']]
h['Tiltaxis'] = extended['tiltaxis'][0]
h['Username'] = None
h['Date'] = None
h['Size'] = [mrc['nx'], mrc['ny'], mrc['nz']]
h['Comment'] = None
h['Parameter'] = None
h['Fillup'] = None
h['Filename'] = path
h['Marker_X'] = None
h['Marker_Y'] = None
h['MRC'] = mrc
return {
'header': h,
'value': v,
}
def __init__(self, fileName):
f = open(fileName, 'rb')
a = np.fromfile(f, dtype=np.uint64)
self.reqTimes = a.reshape((a.shape[0], 1))
f.close()
# set up model
model = Sequential()
model.add(Conv1D(64, 5, input_shape=(128, 44), activation='relu'))
model.add(Dropout(0.2))
model.add(Conv1D(32, 5, activation='relu'))
model.add(Dropout(0.2))
model.add(Flatten())
model.add(Dense(1))
model.compile(optimizer='rmsprop', loss='mse')
print('done setting up model!')
# get training and validation data
X = np.load('data/train_frames.npy')
y = np.fromfile('data/train.txt', sep=' ')[:-1]
idx = np.random.permutation(len(X))
X, y = X[idx], y[idx]
X_validate = X[:len(X) // 5]
X_train = X[len(X) // 5:]
y_validate = y[:len(X) // 5]
y_train = y[len(X) // 5:]
print('done loading data!')
model.fit(X_train, y_train, epochs=40, batch_size=32)
print('done fitting model!')
print(model.summary())
y_validate_pred = model.predict(X_validate)
def read_labels(path_to_labels):
with open(path_to_labels, 'rb') as f:
labels = np.fromfile(f, dtype=np.uint8)
labels = labels.astype(np.uint8)
return labels
def readFromGIAnT(self, h5file, setmaster2zero=None,
zfile=None, lonfile=None, latfile=None, filetype='f',
incidence=None, heading=None, inctype='onefloat',
field='recons', keepnan=False, mask=None, readModel=False):
'''
Read the output from a typical GIAnT h5 output file.
Args:
* h5file : Input h5file
Kwargs:
* setmaster2zero: If index is provided, master will be replaced by zeros (no substraction)
* zfile : File with elevation
* lonfile : File with longitudes
* latfile : File with latitudes
* filetype : type of data in lon, lat and elevation file (default: 'f')
* incidence : Incidence angle (degree)
* heading : Heading angle (degree)
* inctype : Type of the incidence and heading values (see insar.py for details). Can be 'onefloat', 'grd', 'binary', 'binaryfloat'
* field : Name of the field in the h5 file.
* mask : Adds a common mask to the data. mask is an array the same size as the data with nans and 1. It can also be a tuple with a key word in the h5file, a value and 'above' or 'under'
* readModel : Reads the model parameters
Returns:
* None
'''
# open the h5file
h5in = h5py.File(h5file, 'r')
self.h5in = h5in
# Get the data
data = h5in[field]
# Get some sizes
nDates = data.shape[0]
nLines = data.shape[1]
nCols = data.shape[2]
# Deal with the mask instructions
if mask is not None:
if type(mask) is tuple:
key = mask[0]
value = mask[1]
instruction = mask[2]
mask = np.ones((nLines, nCols))
if instruction in ('above'):
mask[np.where(h5in[key][:]>value)] = np.nan
elif instruction in ('under'):
mask[np.where(h5in[key][:]<value)] = np.nan
else:
print('Unknow instruction type for Masking...')
sys.exit(1)
# Read Lon Lat
if lonfile is not None:
self.lon = np.fromfile(lonfile, dtype=filetype)
if latfile is not None:
self.lat = np.fromfile(latfile, dtype=filetype)
# Compute utm
self.x, self.y = self.ll2xy(self.lon, self.lat)
# Elevation
if zfile is not None:
self.elevation = insar('Elevation', utmzone=self.utmzone,
verbose=False, lon0=self.lon0, lat0=self.lat0,
ellps=self.ellps)
self.elevation.read_from_binary(zfile, lonfile, latfile,
incidence=None, heading=None,
remove_nan=False, remove_zeros=False,
dtype=filetype)
self.z = self.elevation.vel
# Get the time
dates = h5in['dates']
self.time = []
for i in range(nDates):
self.time.append(dt.datetime.fromordinal(int(dates[i])))
# Create a list to hold the dates
self.timeseries = []
# Iterate over the dates
for i in range(nDates):
# Get things
date = self.time[i]
dat = data[i,:,:]
# Mask?
if mask is not None:
dat *= mask
# check master date
if i is setmaster2zero:
dat[:,:] = 0.
# Create an insar object
sar = insar('{} {}'.format(self.name,date.isoformat()), utmzone=self.utmzone,
verbose=False, lon0=self.lon0, lat0=self.lat0, ellps=self.ellps)
# Put thing in the insarrate object
sar.vel = dat.flatten()
sar.lon = self.lon
sar.lat = self.lat
sar.x = self.x
sar.y = self.y
# Things should remain None
sar.corner = None
sar.err = None
# Set factor
sar.factor = 1.0
# Take care of the LOS
if incidence is not None and heading is not None:
sar.inchd2los(incidence, heading, origin=inctype)
else:
sar.los = np.zeros((sar.vel.shape[0], 3))
# Store the object in the list
self.timeseries.append(sar)
# Keep incidence and heading
self.incidence = incidence
self.heading = heading
self.inctype = inctype
# if readVel
if readModel:
self.readModelFromGIAnT()
# Make a common mask if asked
if not keepnan:
# Create an array
checkNaNs = np.zeros(self.lon.shape)
checkNaNs[:] = False
# Trash the pixels where there is only NaNs
for sar in self.timeseries:
checkNaNs += np.isfinite(sar.vel)
uu = np.flatnonzero(checkNaNs==0)
# Keep 'em
for sar in self.timeseries:
sar.reject_pixel(uu)
if zfile is not None:
elevation.reject_pixel(uu)
self.reject_pixel(uu)
h5in.close()
# all done
return
"""
Tally benchmarks
"""
import os, glob
import numpy as np
normalize = 0
np_ = []
tt = []
for d in glob.glob( '[0-9]*' ):
prm = {}
path = os.path.join( d, 'parameters.py' )
exec open( path ) in prm
np_ += [ np.product( prm['np3'] ) ]
t = np.fromfile( d + '/prof/8step', 'f' )
tt += [ np.sum( t[1:-1] ) / (len(t)-2) ]
if normalize:
tt = [ t / tt[0] for t in tt ]
print 'time cores'
for n, t in zip( np_, tt ):
print '%4.2f %d' % (t, n)
if 0:
import matplotlib.pyplot as plt
n = len( tt )
ax = plt.plot( tt, 'ko-' )[0].axes
ax.hold( True )
ax.plot( [-1, n], [tt[0],tt[0]], 'k--' )
def _read_cells(f, line):
# If the line is self-contained, it is merely a declaration of the total number of
# points.
if line.count("(") == line.count(")"):
return None, None
out = re.match("\\s*\\(\\s*(|20|30)12\\s*\\(([^\\)]+)\\).*", line)
assert out is not None
a = [int(num, 16) for num in out.group(2).split()]
if len(a) <= 4:
raise ReadError()
first_index = a[1]
last_index = a[2]
num_cells = last_index - first_index + 1
zone_type = a[3]
element_type = a[4]
if zone_type == 0:
# dead zone
return None, None
key, num_nodes_per_cell = {
0: ("mixed", None),
1: ("triangle", 3),
2: ("tetra", 4),
3: ("quad", 4),
4: ("hexahedron", 8),
5: ("pyramid", 5),
6: ("wedge", 6),
}[element_type]
# Skip to the opening `(` and make sure that there's no non-whitespace character
# between the last closing bracket and the `(`.
if line.strip()[-1] != "(":
c = None
while True:
c = f.read(1).decode()
if c == "(":
break
if not re.match("\\s", c):
# Found a non-whitespace character before `(`.
# Assume this is just a declaration line then and
# skip to the closing bracket.
_skip_to(f, ")")
return None, None
if key == "mixed":
# From
# <https://www.afs.enea.it/project/neptunius/docs/fluent/html/ug/node1470.htm>:
#
# > If a zone is of mixed type (element-type=0), it will have a body that
# > lists the element type of each cell.
#
# No idea where the information other than the element types is stored
# though. Skip for now.
data = None
else:
# read cell data
if out.group(1) == "":
# ASCII cells
data = np.empty((num_cells, num_nodes_per_cell), dtype=int)
for k in range(num_cells):
line = f.readline().decode()
dat = line.split()
if len(dat) != num_nodes_per_cell:
raise ReadError()
data[k] = [int(d, 16) for d in dat]
else:
if key == "mixed":
raise ReadError("Cannot read mixed cells in binary mode yet")
# binary cells
if out.group(1) == "20":
dtype = np.int32
else:
if out.group(1) != "30":
ReadError(f"Expected keys '20' or '30', got {out.group(1)}.")
dtype = np.int64
shape = (num_cells, num_nodes_per_cell)
count = shape[0] * shape[1]
data = np.fromfile(f, count=count, dtype=dtype).reshape(shape)
# make sure that the data set is properly closed
_skip_close(f, 2)
return key, data
def session_plot(file_path, fig_no=1, return_fig=False):
'''Plot the states, events and analog data for a pyControl session.'''
# Import data file.
with open(file_path, 'r') as f:
all_lines = [line.strip() for line in f.readlines() if line.strip()]
# Import any analog files.
file_dir = os.path.dirname(file_path)
file_name = os.path.split(file_path)[1]
analog_files = [
f for f in os.listdir(file_dir)
if file_name.split('.')[0] in f and f != file_name
]
analog_data = {}
for analog_file in analog_files:
analog_name = analog_file[len(file_name.split('.')[0]) + 1:-4]
with open(os.path.join(file_dir, analog_file), 'rb') as f:
analog_data[analog_name] = np.fromfile(f,
dtype='<i').reshape(-1, 2)
# Extract state entry and event times.
states_dict = eval(next(line for line in all_lines if line[0] == 'S')[2:])
events_dict = eval(next(line for line in all_lines if line[0] == 'E')[2:])
ID2name = {v: k for k, v in {**states_dict, **events_dict}.items()}
data_lines = [line[2:].split(' ') for line in all_lines if line[0] == 'D']
event_times = np.array([
int(dl[0]) for dl in data_lines if int(dl[1]) in events_dict.values()
]) / 1000
event_IDs = np.array([
int(dl[1]) for dl in data_lines if int(dl[1]) in events_dict.values()
])
state_times = np.array([
int(dl[0]) for dl in data_lines if int(dl[1]) in states_dict.values()
]) / 1000
state_IDs = np.array([
int(dl[1]) for dl in data_lines if int(dl[1]) in states_dict.values()
])
state_durations = np.diff(state_times)
# Plotting
fig = plt.figure(fig_no, figsize=[18, 12])
fig.clf()
n_subplots = 3 if analog_data else 2
# Plot states.
ax1 = plt.subplot(n_subplots, 1, 1)
plt.quiver(state_times[:-1],
state_IDs[:-1],
state_durations,
np.zeros(state_durations.shape),
state_IDs[:-1],
cmap='gist_rainbow',
headwidth=1,
headlength=0,
minlength=0,
scale=1,
width=10,
units='dots',
scale_units='xy')
ax1.set_yticks(sorted(states_dict.values()))
ax1.set_yticklabels([ID2name[ID] for ID in sorted(states_dict.values())])
ax1.set_ylim(
min(states_dict.values()) - 0.5,
max(states_dict.values()) + 0.5)
ax1.set_facecolor('black')
# Plot events.
ax2 = plt.subplot(n_subplots, 1, 2, sharex=ax1)
plt.scatter(event_times, event_IDs, c=event_IDs, s=6, cmap='gist_rainbow')
ax2.set_yticks(sorted(events_dict.values()))
ax2.set_yticklabels([ID2name[ID] for ID in sorted(events_dict.values())])
ax2.set_ylim(
min(events_dict.values()) - 0.5,
max(events_dict.values()) + 0.5)
ax2.set_facecolor('black')
# Plot analog data
if analog_data:
ax3 = plt.subplot(n_subplots, 1, 3, sharex=ax1)
for name, data in analog_data.items():
plt.plot(data[:, 0] / 1000, data[:, 1], label=name)
ax3.set_facecolor('black')
ax3.set_ylabel('Signal value')
ax3.legend()
ax1.set_xlim(0, state_times[-1])
plt.xlabel('Time (seconds)')
plt.tight_layout()
if return_fig: # Return the figure and axes
return fig, ax1
