def render_line2(filename, frame_rate=60):
if is_animation(parse_tags(filename)):
data = []
def get_line_data(filename, frame):
filename = filename.replace('0000', '%04d' % frame)
data_x = np.load(filename)
data_y = np.load(filename.replace(',' + X_TAG, ',' + Y_TAG))
return (data_x, data_y)
def update_lines(frame, line):
x, y = data[frame]
line.set_data(x, y)
return line,
dirname = os.path.dirname(filename)
basename = os.path.basename(filename).replace('0000', '[0-9][0-9][0-9][0-9]')
seq = utils.get_all_files(dirname, [basename])
seq.sort()
if len(seq) == 0:
return
data_x, data_y = get_line_data(seq[0], 0)
has_bbox = False
for frame in range(len(seq)):
data_x, data_y = get_line_data(seq[frame], frame)
if len(data_x) > 0 and len(data_y) > 0:
if has_bbox:
xmin = min(xmin, data_x.min())
xmax = max(xmax, data_x.max())
ymin = min(ymin, data_y.min())
ymax = max(ymax, data_y.max())
else:
xmin = data_x.min()
xmax = data_x.max()
ymin = data_y.min()
ymax = data_y.max()
has_bbox = True
data.append((data_x, data_y))
fig, ax = plt.subplots()
x, y = get_line_data(filename, 0)
if has_bbox:
xmid = (xmax + xmin) / 2.0
ymid = (ymax + ymin) / 2.0
new_xmin = xmid - 1.25 * (xmid - xmin)
new_xmax = xmid + 1.25 * (xmax - xmid)
new_ymin = ymid - 1.25 * (ymid - ymin)
new_ymax = ymid + 1.25 * (ymax - ymid)
ax.set_xlim(new_xmin, new_xmax)
ax.set_ylim(new_ymin, new_ymax)
ax.set_aspect('equal')
line, = ax.plot(x, y, lw=2, marker='o', markersize=3)
anim = animation.FuncAnimation(fig, update_lines, len(seq), fargs=(line,), interval=60, blit=False)
output_filename = get_output_movie_filename(filename.replace(',' + X_TAG, ''))
anim.save(output_filename, fps=frame_rate, bitrate=5000, writer=video_writer, extra_args=video_extra_args)
plt.close(fig)
print ('Rendered <%s>' % output_filename)
else:
output_filename = get_output_bitmap_filename(filename.replace(',' + X_TAG, ''))
render_still_line2(filename, filename.replace(',' + X_TAG, ',' + Y_TAG), output_filename)
def read_action(action_path):
action_name=get_action_name(action_path)
category=get_category(action_name)
person=get_person(action_name)
all_files=utils.get_all_files(action_path)
all_files=utils.append_path(action_path+"/",all_files)
frames=utils.read_images(all_files)
return Action(category,person,frames)
def load_data(path,batch_size=25):
all_files=utils.get_all_files(path)
all_files=utils.append_path(path,all_files)
images=utils.read_images(all_files)
images=utils.flatten_images(images)
images=map(utils.normalize,images)
images=np.array(images)
n_batches=get_number_of_batches(batch_size,len(images))
def get_batch(i):
return images[i * batch_size: (i+1) * batch_size]
batches=map(get_batch,range(n_batches))
batches = [np.array(batch) for batch in batches]
print("Dataset loaded")
return np.array(batches)
def main():
"""Hook spot for the console script."""
parser = optparse.OptionParser()
parser.add_option('-p', '--paths', action='store_true',
help='Preserve path structure of original files')
parser.add_option('-d', '--directory', action='store', type='string',
dest='outdir', default='.',
help='The output directory that the rendered files should go to.')
parser.add_option('-w', '--watch', action='store_true',
help='Watch original files and re-generate documentation on changes')
parser.add_option('-a', '--all', action='store_true',
help='Get all files from subfolders')
opts, sources = parser.parse_args()
if not sources:
return
filepath = os.path.dirname(sources[0])
start, filetype = os.path.splitext(sources[0])
if start.endswith('*'):
return
start = os.path.dirname(os.path.dirname(os.path.abspath(start)))
if opts.all:
sources = [i for i in get_all_files(filepath or '.', filetype)]
global SOURCES
SOURCES = sorted([Source(name, start) for name in sources])
process(SOURCES, outdir=opts.outdir, preserve_paths=opts.paths)
# If the -w / --watch option was present, monitor the source directories
# for changes and re-generate documentation for source files whenever they
# are modified.
if opts.watch:
try:
import watchdog.events
import watchdog.observers
except ImportError:
sys.exit('The -w/--watch option requires the watchdog package.')
monitor(sources, opts)
def SPVideoFromSequence(context):
args = context.args
if args.referenceFolder:
referenceFile = get_all_files(args.referenceFolder)
VideoEd.video_from_sequence(
args.inputData,
os.path.join(args.outputData, args.fileName),
referenceFile[0],
args.ext,
args.fps,
args.bitrate,
args.lossless,
)
return args.outputData
def main(argv=None):
args = docopt(__doc__, argv=argv)
pictures = utils.get_all_files(args['<input>'])
destination = args['<destination>']
dry_run = args['--dry-run']
move = args['--move']
verbose = args['--verbose']
groups = group_pictures(pictures)
groups = group_consecutive_days(groups)
for group, pics in groups.iteritems():
print group, len(pics)
send_to_destination(destination, groups, move=move, dry_run=dry_run, verbose=verbose)
def bbai(dirs, outfile='./data_info', model=''):
files = utils.get_all_files(dirs)
if model == 'fametrecafe':
model = '/home/pgrespan/trained/VGG16_2020-04-21_11-01/VGG16_13_0.88753_0.86502.h5'
gen = LSTGenerator(h5files=files,
feature="energy",
shuffle=False,
intensity=None,
leakage2_intensity=None,
class_model=model)
try:
gen.get_all_info().to_pickle(outfile + '.pkl')
except:
print('No riesso catare a cartea, probabilmente no esiste, molton!')
return gen.indexes
def test_classifiers():
model_dir= 'models/'
test_file = 'feature/clean_features.csv'
result_dir = 'result/'
result_file = 'result.csv'
model_files = get_all_files(model_dir)
for mfile in model_files:
model_file = os.path.join(model_dir,mfile) #get the path of the model file
with open(model_file, 'rb') as fp:
trained_model = cPickle.load(fp) # load trained model
data = np.genfromtxt(test_file, delimiter=',') # read the test file
row, col = data.shape
x = data[:, 1:col - 1] # get data without the label information
y_true = data[:, col - 1] # get the label information
y_pred = trained_model.predict(x) # perform prediction
acc = accuracy_score(y_true, y_pred)
f1 = f1_score(y_true, y_pred, average=None)
prc = precision_score(y_true, y_pred, average=None)
rcl = recall_score(y_true, y_pred, average=None)
if not os.path.exists(result_dir):
os.makedirs(result_dir)
res_file = os.path.join(result_dir,result_file)
with open(res_file,'w') as fp:
print 'writing individual results...'
fp.write("acc :: " + str(acc))
fp.write('\n\n')
fp.write('f1 :: ' + ','.join(str(x) for x in f1))
fp.write('\n\n')
fp.write("precision :: " + ','.join(
str(x) for x in prc))
fp.write('\n\n')
fp.write("recall :: " + ','.join(str(x) for x in rcl))
fp.write('\n\n\n')
def main():
include_dir = os.path.join(dirname, '../include/jet')
filenames = utils.get_all_files(include_dir, ['*.h'])
filenames.sort()
header = os.path.join(dirname, '../include/jet/jet.h')
with open(header, 'w') as header_file:
header_file.write("""// Copyright (c) 2017 Doyub Kim
// I am making my contributions/submissions to this project solely in my
// personal capacity and am not conveying any rights to any intellectual
// property of any third parties.\n
""")
header_file.write('#ifndef INCLUDE_JET_JET_H_\n')
header_file.write('#define INCLUDE_JET_JET_H_\n')
for filename in filenames:
if not filename.endswith('-inl.h'):
line = '#include <jet/%s>\n' % os.path.basename(filename)
header_file.write(line)
header_file.write('#endif // INCLUDE_JET_JET_H_\n')
def main():
include_dir = os.path.join(dirname, '../include/jet')
filenames = utils.get_all_files(include_dir, ['*.h'])
filenames.sort()
header = os.path.join(dirname, '../include/jet/jet.h')
with open(header, 'w') as header_file:
header_file.write("""// Copyright (c) 2017 Doyub Kim
// I am making my contributions/submissions to this project solely in my
// personal capacity and am not conveying any rights to any intellectual
// property of any third parties.\n
""")
header_file.write('#ifndef INCLUDE_JET_JET_H_\n')
header_file.write('#define INCLUDE_JET_JET_H_\n')
for filename in filenames:
if not filename.endswith('-inl.h'):
line = '#include <jet/%s>\n' % os.path.basename(filename)
header_file.write(line)
header_file.write('#endif // INCLUDE_JET_JET_H_\n')
def __init__(self,
batch_size,
num_vid_frames=4,
framerate=30,
framesize=256,
samplerate=16000,
max_vid_frames=100,
noise_std=0.01,
center_fft=True,
use_polar=True,
shuffle_files=True,
data_path="./data/raw"):
assert batch_size > 1
self.attention_extractor = VideoAttention()
self.max_vid_frames = max_vid_frames
self.batch_size = batch_size
self.num_vid_frames = num_vid_frames
self.framerate = framerate
self.framesize = framesize
self.samplerate = samplerate
self.noise_std = noise_std
self.normalize_input_fft = False
self.fft_len = int(
(self.num_vid_frames / self.framerate) * self.samplerate)
self.all_vids = utils.get_all_files(data_path, "mp4")
print(f'number of videos found: {len(self.all_vids)}')
if shuffle_files:
random.shuffle(self.all_vids)
# self.current_av_pair = self.load_example_pair(self.all_vids[0])
self.example_idx = 0
self.center_fft = center_fft
self.use_polar = use_polar
def main(argv=None):
args = docopt(__doc__, argv=argv)
pictures = utils.get_all_files(args['<input>'])
destination = args['<destination>']
dry_run = args['--dry-run']
move = args['--move']
verbose = args['--verbose']
groups = group_pictures(pictures)
groups = group_consecutive_days(groups)
for group, pics in groups.iteritems():
print group, len(pics)
send_to_destination(destination,
groups,
move=move,
dry_run=dry_run,
verbose=verbose)
def main():
include_dir = os.path.join(dirname, "../include/blending")
filenames = utils.get_all_files(include_dir, ["*.h"])
filenames.sort()
header = os.path.join(dirname, "../include/blending/blending.h")
header_tmp = header + ".tmp"
with open(header_tmp, "w") as header_file:
header_file.write("""// Copyright (c) 2018 loyave
""")
header_file.write("#ifndef INCLUDE_BLENDING_BLENDING_H_\n")
header_file.write("#define INCLUDE_BLENDING_BLENDING_H_\n")
for filename in filenames:
if not filename.endswith("-inl.h"):
line = "#include <blending/%s>\n" % os.path.basename(filename)
header_file.write(line)
header_file.write("#endif // INCLUDE_BLENDING_BLENDING_H_\n")
if not filecmp.cmp(header, header_tmp):
shutil.move(header_tmp, header)
else:
os.remove(header_tmp)
def process_all(root, output):
assert formatted or compact
src_files = get_all_files(root, '.json')
for f in tqdm(src_files):
game = json.load(open(f, 'r'))
game = process_game(game, f)
dest_file = f.replace(root, output)
dirname = os.path.dirname(dest_file)
if not os.path.exists(dirname):
os.makedirs(dirname)
temp_json = tempfile.TemporaryFile('r+')
json.dump(game, temp_json, indent=4)
temp_json.seek(0)
formatted = format_json(temp_json)
with open(dest_file, 'w') as f:
f.write(formatted)
def main(input, output, emin, emax, log_energy, its, lkg):
files = get_all_files(input)
print("Indexing data...")
generator = DataGeneratorR(files,
feature="energy",
shuffle=False,
intensity=its,
emin=emin,
emax=emax)
print("Loading data and printing images...")
title = "Gamma"
for i in trange(generator.__len__()):
images, energies = generator.__getitem__(i)
#print("Images shape: ", images.shape)
save_interp_images(charges=images,
title=title,
energies=energies,
start=i,
outputdir=output)
print("Finished!")
def render_grid2(filename, frame_rate=60):
if is_animation(parse_tags(filename)):
dirname = os.path.dirname(filename)
basename = os.path.basename(filename).replace('0000', '[0-9][0-9][0-9][0-9]')
seq = utils.get_all_files(dirname, [basename])
seq.sort()
if len(seq) == 0:
return
grid_data = np.load(seq[0])
grid_data = np.flipud(grid_data)
fig, ax = plt.subplots()
im = ax.imshow(grid_data, cmap=plt.cm.gray, interpolation='nearest')
if ISO_TAG in parse_tags(filename):
plt.contour(grid_data)
def update_image(i):
grid_data = np.load(seq[i])
grid_data = np.flipud(grid_data)
im.set_array(grid_data)
if ISO_TAG in parse_tags(filename):
plt.contour(grid_data)
return im
output_filename = get_output_movie_filename(filename.replace(',' + X_TAG, ''))
anim = animation.FuncAnimation(fig, update_image, frames=len(seq), interval=60, blit=False)
anim.save(output_filename, fps=frame_rate, bitrate=5000, writer=video_writer, extra_args=video_extra_args)
plt.close(fig)
print ('Rendered <%s>' % output_filename)
else:
tags = parse_tags(filename)
output_filename = get_output_bitmap_filename(filename)
if X_TAG in tags:
grid_data_u = np.load(filename)
grid_data_v = np.load(filename.replace(X_TAG, Y_TAG))
render_still_vector_grid2(filename, filename.replace(X_TAG, Y_TAG), output_filename)
else:
has_iso = ISO_TAG in parse_tags(filename)
render_still_scalar_grid2(filename, output_filename, has_iso=has_iso)
def SPConvert(context):
args = context.args
video = get_all_files(args.refVideo)[0] if args.refVideo else args.refVideo
PARSER = cli.FullHelpArgumentParser()
CONVERT = cli.ConvertArgs(
PARSER, "convert",
"Convert a source image to a new one with the face swapped")
argsTransfer = [
"--input-dir",
args.inputData1,
"--model-dir",
args.inputModel,
"--output-dir",
args.outputData,
"--color-adjustment",
args.colorAdjustment,
"--mask-type",
args.maskType,
"--scaling",
args.scaling,
"--gpus",
str(args.__gpu),
]
if args.inputData2:
argsTransfer += [
"--alignments",
os.path.join(args.inputData2, "alignments.json"),
]
ARGUMENTS = PARSER.parse_args(argsTransfer)
ARGUMENTS.func(ARGUMENTS)
return args.outputData
def tester(folders, mdl, batch_size, atime, workers, intensity):
h5files = get_all_files(folders)
random.shuffle(h5files)
model = load_model(mdl)
print('Building test generator...')
test_generator = DataGeneratorC(h5files,
batch_size=batch_size,
arrival_time=atime,
shuffle=False,
intensity=intensity)
print('Number of test batches: ' + str(len(test_generator)))
pr_labels = model.predict_generator(generator=test_generator,
steps=None,
max_queue_size=10,
workers=workers,
use_multiprocessing=True,
verbose=1)
test_idxs = test_generator.get_indexes() # ground truth labels
gt_labels = test_idxs[:, 2].reshape(test_idxs[:, 2].shape[0])
pr_labels = pr_labels.reshape(pr_labels.shape[0])
# saving predictions
df = pd.DataFrame()
df['GroundTruth'] = gt_labels
df['Predicted'] = pr_labels
df.to_csv(mdl + '_test.csv', sep=',', index=False, encoding='utf-8')
p_file = mdl + '_test.csv'
print('Results saved in ' + p_file)
return p_file
def main(annotation_file, image_dir, chip_dir, output_dir, show_ggnn, infer,
_run):
refiner = PolygonRefiner(infer=infer, show_ggnn=show_ggnn)
if os.path.isfile(annotation_file):
refiner.process_file(
annotation_file,
image_dir,
chip_dir,
output_dir,
)
else:
image_dirs = get_all_files(annotation_file, isdir=True, extension="*")
for image_dir in image_dirs:
(child_chip_dir, child_output_dir) = [
os.path.join(x, os.path.basename(image_dir))
for x in (chip_dir, output_dir)
]
annotation_file = os.path.join(image_dir, "annotations.csv")
refiner.process_file(
annotation_file,
image_dir,
child_chip_dir,
child_output_dir,
)
def main():
include_dir = os.path.join(dirname, "../include/jet")
filenames = utils.get_all_files(include_dir, ["*.h"])
filenames.sort()
header = os.path.join(dirname, "../include/jet/jet.h")
header_tmp = header + ".tmp"
with open(header_tmp, "w") as header_file:
header_file.write("""// Copyright (c) 2018 Doyub Kim
// I am making my contributions/submissions to this project solely in my
// personal capacity and am not conveying any rights to any intellectual
// property of any third parties.\n
""")
header_file.write("#ifndef INCLUDE_JET_JET_H_\n")
header_file.write("#define INCLUDE_JET_JET_H_\n")
for filename in filenames:
if not filename.endswith("-inl.h"):
line = "#include <jet/%s>\n" % os.path.basename(filename)
header_file.write(line)
header_file.write("#endif // INCLUDE_JET_JET_H_\n")
if not filecmp.cmp(header, header_tmp):
shutil.move(header_tmp, header)
else:
os.remove(header_tmp)
def main():
include_dir = os.path.join(dirname, "../include/jet")
filenames = utils.get_all_files(include_dir, ["*.h"])
filenames.sort()
header = os.path.join(dirname, "../include/jet/jet.h")
header_tmp = header + ".tmp"
with open(header_tmp, "w") as header_file:
header_file.write("""// Copyright (c) 2017 Doyub Kim
// I am making my contributions/submissions to this project solely in my
// personal capacity and am not conveying any rights to any intellectual
// property of any third parties.\n
""")
header_file.write("#ifndef INCLUDE_JET_JET_H_\n")
header_file.write("#define INCLUDE_JET_JET_H_\n")
for filename in filenames:
if not filename.endswith("-inl.h"):
line = "#include <jet/%s>\n" % os.path.basename(filename)
header_file.write(line)
header_file.write("#endif // INCLUDE_JET_JET_H_\n")
if not filecmp.cmp(header, header_tmp):
shutil.move(header_tmp, header)
else:
os.remove(header_tmp)
_x = []
_y = []
for xyz in xyz_file:
_x.append(float(xyz.split()[0]))
_y.append(float(xyz.split()[1]))
line.set_data(_x, _y)
xyz_file.close()
return line,
# Simulation Path
path = sys.argv[1]
output_name = sys.argv[2] if len(sys.argv) > 2 else 'result.mp4'
files = utils.get_all_files(path, "*.xyz")
size_files = len(files)
# Set up formatting for the movie files
Writer = animation.writers['ffmpeg']
writer = Writer(fps=25, bitrate=1800)
fig1 = plt.figure()
data = []
l, = plt.plot([], [], 'ro')
plt.xlim(0, 1)
plt.ylim(0, 1)
plt.xlabel('x')
plt.title('Simulation Result')
def main():
global _threadError
num_threads = os.cpu_count() * 2
root_dir = path.join(utils.get_script_folder(), '..')
docs_dir = path.join(root_dir, 'docs')
xml_dir = path.join(docs_dir, 'xml')
html_dir = path.join(docs_dir, 'html')
mcss_dir = path.join(root_dir, 'extern', 'mcss')
doxygen = path.join(mcss_dir, 'documentation', 'doxygen.py')
# delete any leftovers from the previous run
if 1:
utils.delete_directory(xml_dir)
utils.delete_directory(html_dir)
# run doxygen to generate the xml
if 1:
subprocess.check_call(['doxygen', 'Doxyfile'],
shell=True,
cwd=docs_dir)
# fix some shit that's broken in the xml
if 1:
preprocess_xml(xml_dir)
# run doxygen.py (m.css) to generate the html
if 1:
utils.run_python_script(doxygen, path.join(docs_dir, 'Doxyfile-mcss'),
'--no-doxygen')
# post-process html files
if 1:
fixes = [
CustomTagsFix(),
SyntaxHighlightingFix(),
IndexPageFix(),
ModifiersFix1(),
ModifiersFix2(),
ExtDocLinksFix(),
EnableIfFix(),
ExternalLinksFix(),
TemplateTemplateFix()
]
files = [
path.split(f)
for f in utils.get_all_files(html_dir, any=('*.html', '*.htm'))
]
if files:
with futures.ThreadPoolExecutor(
max_workers=min(len(files), num_threads)) as executor:
jobs = {
executor.submit(postprocess_file, dir, file, fixes): file
for dir, file in files
}
for job in futures.as_completed(jobs):
if _threadError:
executor.shutdown(False)
break
else:
file = jobs[job]
print('Finished processing {}.'.format(file))
if _threadError:
return 1
def preprocess_xml(dir):
global inline_namespaces
global implementation_headers
if not inline_namespaces and not implementation_headers:
return
write_xml_to_file = lambda x, f: x.write(
f, encoding='utf-8', xml_declaration=True)
inline_namespace_ids = [
f'namespace{doxygen_mangle_name(ns)}' for ns in inline_namespaces
]
implementation_header_data = [
(hp, path.basename(hp), doxygen_mangle_name(path.basename(hp)),
[(i, path.basename(i), doxygen_mangle_name(path.basename(i)))
for i in impl]) for hp, impl in implementation_headers
]
implementation_header_mappings = dict()
implementation_header_innernamespaces = dict()
implementation_header_sectiondefs = dict()
for hdata in implementation_header_data:
implementation_header_innernamespaces[hdata[2]] = []
implementation_header_sectiondefs[hdata[2]] = []
for (ip, ifn, iid) in hdata[3]:
implementation_header_mappings[iid] = hdata
if 1:
extracted_implementation = False
xml_files = utils.get_all_files(dir, any=('*.xml'))
for xml_file in xml_files:
print(f'Pre-processing {xml_file}')
xml = ET.parse(xml_file)
root = xml.getroot().find('compounddef')
if root is None:
continue
changed = False
# namespaces
if root.get("kind") == "namespace" and inline_namespaces:
for nsid in inline_namespace_ids:
if root.get("id") == nsid:
root.set("inline", "yes")
changed = True
break
# dirs
if root.get("kind") == "dir" and implementation_headers:
innerfiles = root.findall('innerfile')
for innerfile in innerfiles:
if innerfile.get(
'refid') in implementation_header_mappings:
root.remove(innerfile)
changed = True
# header files
if root.get("kind") == "file" and implementation_headers:
# remove junk not required by m.css
for tag in ('includes', 'includedby', 'incdepgraph',
'invincdepgraph'):
tags = root.findall(tag)
if tags:
for t in tags:
root.remove(t)
changed = True
# rip the good bits out of implementation headers
if root.get("id") in implementation_header_mappings:
hid = implementation_header_mappings[root.get("id")][2]
innernamespaces = root.findall('innernamespace')
if innernamespaces:
implementation_header_innernamespaces[
hid] = implementation_header_innernamespaces[
hid] + innernamespaces
extracted_implementation = True
for tag in innernamespaces:
root.remove(tag)
changed = True
sectiondefs = root.findall('sectiondef')
if sectiondefs:
implementation_header_sectiondefs[
hid] = implementation_header_sectiondefs[
hid] + sectiondefs
extracted_implementation = True
for tag in sectiondefs:
root.remove(tag)
changed = True
if changed:
write_xml_to_file(xml, xml_file)
# merge extracted implementations
if extracted_implementation:
for (hp, hfn, hid, impl) in implementation_header_data:
xml_file = path.join(dir, f'{hid}.xml')
print(f'Merging implementation nodes into {xml_file}')
xml = ET.parse(xml_file)
root = xml.getroot().find('compounddef')
changed = False
innernamespaces = root.findall('innernamespace')
for new_tag in implementation_header_innernamespaces[hid]:
matched = False
for existing_tag in innernamespaces:
if existing_tag.get('refid') == new_tag.get('refid'):
matched = True
break
if not matched:
root.append(new_tag)
innernamespaces.append(new_tag)
changed = True
sectiondefs = root.findall('sectiondef')
for new_section in implementation_header_sectiondefs[hid]:
matched_section = False
for existing_section in sectiondefs:
if existing_section.get('kind') == new_section.get(
'kind'):
matched_section = True
memberdefs = existing_section.findall('memberdef')
new_memberdefs = new_section.findall('memberdef')
for new_memberdef in new_memberdefs:
matched = False
for existing_memberdef in memberdefs:
if existing_memberdef.get(
'id') == new_memberdef.get('id'):
matched = True
break
if not matched:
new_section.remove(new_memberdef)
existing_section.append(new_memberdef)
memberdefs.append(new_memberdef)
changed = True
break
if not matched_section:
root.append(new_section)
sectiondefs.append(new_section)
changed = True
if changed:
write_xml_to_file(xml, xml_file)
# delete the impl header xml files
if 1 and implementation_headers:
for hdata in implementation_header_data:
for (ip, ifn, iid) in hdata[3]:
xml_file = path.join(dir, f'{iid}.xml')
if (path.exists(xml_file)):
print(f'Deleting {xml_file}')
os.remove(xml_file)
# scan through the files and substitute impl header ids and paths as appropriate
if 1 and implementation_headers:
xml_files = utils.get_all_files(dir, any=('*.xml'))
for xml_file in xml_files:
print(f'Re-linking implementation headers in {xml_file}')
xml = utils.read_all_text_from_file(xml_file)
for (hp, hfn, hid, impl) in implementation_header_data:
for (ip, ifn, iid) in impl:
#xml = xml.replace(f'refid="{iid}"',f'refid="{hid}"')
xml = xml.replace(f'compoundref="{iid}"',
f'compoundref="{hid}"')
xml = xml.replace(ip, hp)
with open(xml_file, 'w', encoding='utf-8', newline='\n') as f:
f.write(xml)
from utils import get_all_files
paths = get_all_files('A')
with open('dir.txt', 'w') as f:
for i in paths:
f.write(i + "\n")
def regressor_training_main(folders,
val_folders,
load_model,
model_name,
time,
epochs,
batch_size,
opt,
learning_rate,
lropf,
sd,
es,
feature,
workers,
test_dirs,
intensity_cut,
leakage,
tb,
gpu_fraction=0.5,
emin=-100,
emax=100,
class_model='',
clr=False,
train_indexes=None,
valid_indexes=None,
clr_values=[5e-5, 5e-3, 4]):
###################################
#TensorFlow wizardry for GPU dynamic memory allocation
if 0 < gpu_fraction < 1:
config = tf.ConfigProto()
# Don't pre-allocate memory; allocate as-needed
config.gpu_options.allow_growth = True
# Only allow a fraction of the GPU memory to be allocated
config.gpu_options.per_process_gpu_memory_fraction = gpu_fraction
# Create a session with the above options specified.
K.tensorflow_backend.set_session(tf.Session(config=config))
###################################
# remove semaphore warnings
os.environ["PYTHONWARNINGS"] = "ignore:semaphore_tracker:UserWarning"
# avoid validation deadlock problem
mp.set_start_method('spawn', force=True)
# hard coded parameters
shuffle = True
channels = 1
if time:
channels = 2
# early stopping
md_es = 0.1 # min delta
p_es = 50 # patience
# cycle learning rate CLR
base_lr = clr_values[0]
max_lr = clr_values[1]
step_size = clr_values[2]
# sgd
lr = 0.01 # lr
decay = 1e-4 # decay
momentum = 0.9 # momentum
nesterov = True
# adam
a_lr = learning_rate
a_beta_1 = 0.9
a_beta_2 = 0.999
a_epsilon = None
a_decay = 0
amsgrad = True
# adabound
ab_lr = 1e-03
ab_final_lr = 0.1
ab_gamma = 1e-03
ab_weight_decay = 0
amsbound = False
# rmsprop
r_lr = 0.01
r_rho = 0.9
r_epsilon = None
r_decay = 0.0
# reduce lr on plateau
f_lrop = 0.1 # factor
p_lrop = 15 # patience
md_lrop = 0.005 # min delta
cd_lrop = 5 # cool down
mlr_lrop = a_lr / 100 # min lr
# cuts
# intensity_cut = 50
leakage2_intensity_cut = leakage
training_files = get_all_files(folders)
validation_files = get_all_files(val_folders)
# create a folder to keep model & results
now = datetime.datetime.now()
root_dir = now.strftime(model_name + '_' + feature + '_' +
'%Y-%m-%d_%H-%M')
mkdir(root_dir)
models_dir = join(root_dir, "models")
mkdir(models_dir)
# generators
print('Building training generator...')
training_generator = LSTGenerator(
training_files,
batch_size=batch_size,
arrival_time=time,
feature=feature,
shuffle=shuffle,
intensity=intensity_cut,
leakage2_intensity=leakage2_intensity_cut,
load_indexes=train_indexes)
if train_indexes is None:
train_idxs = training_generator.get_all_info()
train_idxs.to_pickle(join(root_dir, "train_indexes.pkl"))
if len(val_folders) > 0:
print('Building validation generator...')
validation_generator = LSTGenerator(
validation_files,
batch_size=batch_size,
arrival_time=time,
feature=feature,
shuffle=False,
intensity=intensity_cut,
leakage2_intensity=leakage2_intensity_cut,
load_indexes=valid_indexes)
if valid_indexes is None:
valid_idxs = validation_generator.get_all_info()
valid_idxs.to_pickle(join(root_dir, "valid_indexes.pkl"))
# class_weight = {0: 1., 1: train_protons/train_gammas}
# print(class_weight)
# get image size (rows and columns)
img_rows = training_generator.img_rows
img_cols = training_generator.img_cols
hype_print = '\n' + '======================================HYPERPARAMETERS======================================'
hype_print += '\n' + 'Image rows: ' + str(
img_rows) + ' Image cols: ' + str(img_cols)
hype_print += '\n' + 'Folders:' + str(folders)
hype_print += '\n' + 'Model: ' + str(model_name)
hype_print += '\n' + 'Use arrival time: ' + str(time)
hype_print += '\n' + 'Epochs:' + str(epochs)
hype_print += '\n' + 'Batch size: ' + str(batch_size)
hype_print += '\n' + 'Optimizer: ' + str(opt)
hype_print += '\n' + 'Feature: ' + str(feature)
hype_print += '\n' + 'Validation: ' + str(val_folders)
hype_print += '\n' + 'Test dirs: ' + str(test_dirs)
hype_print += '\n' + 'intensity_cut: ' + str(intensity_cut)
hype_print += '\n' + 'leakage2_intensity_cut: ' + str(
leakage2_intensity_cut)
if clr:
hype_print += '\n' + '--- Cycle Learning Rate ---'
hype_print += '\n' + 'Base LR: ' + str(base_lr)
hype_print += '\n' + 'Max LR: ' + str(max_lr)
hype_print += '\n' + 'Step size: ' + str(step_size) + ' (' + str(
step_size * len(training_generator)) + ')'
if es:
hype_print += '\n' + '--- Early stopping ---'
hype_print += '\n' + 'Min delta: ' + str(md_es)
hype_print += '\n' + 'Patience: ' + str(p_es)
hype_print += '\n' + '----------------------'
if opt == 'sgd':
hype_print += '\n' + '--- SGD ---'
hype_print += '\n' + 'Learning rate:' + str(lr)
hype_print += '\n' + 'Decay: ' + str(decay)
hype_print += '\n' + 'Momentum: ' + str(momentum)
hype_print += '\n' + 'Nesterov: ' + str(nesterov)
hype_print += '\n' + '-----------'
elif opt == 'adam':
hype_print += '\n' + '--- ADAM ---'
hype_print += '\n' + 'lr: ' + str(a_lr)
hype_print += '\n' + 'beta_1: ' + str(a_beta_1)
hype_print += '\n' + 'beta_2: ' + str(a_beta_2)
hype_print += '\n' + 'epsilon: ' + str(a_epsilon)
hype_print += '\n' + 'decay: ' + str(a_decay)
hype_print += '\n' + 'Amsgrad: ' + str(amsgrad)
hype_print += '\n' + '------------'
elif opt == 'rmsprop':
hype_print += '\n' + '--- RMSprop ---'
hype_print += '\n' + 'lr: ' + str(r_lr)
hype_print += '\n' + 'rho: ' + str(r_rho)
hype_print += '\n' + 'epsilon: ' + str(r_epsilon)
hype_print += '\n' + 'decay: ' + str(r_decay)
hype_print += '\n' + '------------'
if lropf:
hype_print += '\n' + '--- Reduce lr on plateau ---'
hype_print += '\n' + 'lr decrease factor: ' + str(f_lrop)
hype_print += '\n' + 'Patience: ' + str(p_lrop)
hype_print += '\n' + 'Min delta: ' + str(md_lrop)
hype_print += '\n' + 'Cool down:' + str(cd_lrop)
hype_print += '\n' + 'Min lr: ' + str(mlr_lrop)
hype_print += '\n' + '----------------------------'
if sd:
hype_print += '\n' + '--- Step decay ---'
hype_print += '\n' + 'Workers: ' + str(workers)
hype_print += '\n' + 'Shuffle: ' + str(shuffle)
hype_print += '\n' + 'Number of training batches: ' + str(
len(training_generator))
if len(val_folders) > 0:
hype_print += '\n' + 'Number of validation batches: ' + str(
len(validation_generator))
outcomes = 1
# loss = 'mean_absolute_percentage_error'
loss = 'mean_absolute_error'
# loss = 'mean_squared_error'
if feature == 'direction':
outcomes = 2
# loss = 'mean_absolute_error'
# loss = 'mean_squared_error'
# keras.backend.set_image_data_format('channels_first')
if load_model:
model = keras.models.load_model(model_name)
model_name = Path(model_name).name
else:
model, hype_print = regressor_selector(model_name, hype_print,
channels, img_rows, img_cols,
outcomes)
hype_print += '\n' + '========================================================================================='
# printing on screen hyperparameters
print(hype_print)
# writing hyperparameters on file
f = open(root_dir + '/hyperparameters.txt', 'w')
f.write(hype_print)
f.close()
model.summary()
callbacks = []
if len(val_folders) > 0:
checkpoint = ModelCheckpoint(
filepath=models_dir + '/' + model_name +
'_{epoch:02d}_{loss:.5f}_{val_loss:.5f}.h5',
monitor='val_loss',
save_best_only=False)
else:
checkpoint = ModelCheckpoint(filepath=models_dir + '/' + model_name +
'_{epoch:02d}_{loss:.5f}.h5',
monitor='loss',
save_best_only=True)
callbacks.append(checkpoint)
# tensorboard = keras.callbacks.TensorBoard(log_dir=root_dir + "/logs",
# histogram_freq=5,
# batch_size=batch_size,
# write_images=True,
# update_freq=batch_size * 100)
history = LossHistoryR()
csv_callback = keras.callbacks.CSVLogger(root_dir + '/epochs_log.csv',
separator=',',
append=False)
callbacks.append(history)
callbacks.append(csv_callback)
# callbacks.append(tensorboard)
# sgd
optimizer = None
if opt == 'sgd':
sgd = optimizers.SGD(lr=lr,
decay=decay,
momentum=momentum,
nesterov=nesterov)
optimizer = sgd
elif opt == 'adam':
adam = optimizers.Adam(lr=a_lr,
beta_1=a_beta_1,
beta_2=a_beta_2,
epsilon=a_epsilon,
decay=a_decay,
amsgrad=amsgrad)
optimizer = adam
# elif opt == 'adabound':
# adabound = AdaBound(lr=ab_lr, final_lr=ab_final_lr, gamma=ab_gamma, weight_decay=ab_weight_decay,
# amsbound=False)
# optimizer = adabound
elif opt == 'rmsprop':
rmsprop = optimizers.RMSprop(lr=r_lr,
rho=r_rho,
epsilon=r_epsilon,
decay=r_decay)
optimizer = rmsprop
# reduce lr on plateau
if lropf:
lrop = keras.callbacks.ReduceLROnPlateau(monitor='val_loss',
factor=f_lrop,
patience=p_lrop,
verbose=1,
mode='auto',
min_delta=md_lrop,
cooldown=cd_lrop,
min_lr=mlr_lrop)
callbacks.append(lrop)
if sd:
# learning rate schedule
def step_decay(epoch):
current = K.eval(model.optimizer.lr)
lrate = current
if epoch == 99:
lrate = current / 10
print('Reduced learning rate by a factor 10')
return lrate
stepd = LearningRateScheduler(step_decay)
callbacks.append(stepd)
if es:
# early stopping
early_stopping = EarlyStopping(monitor='val_loss',
min_delta=md_es,
patience=p_es,
verbose=1,
mode='max')
callbacks.append(early_stopping)
if tb:
tb_path = os.path.join(root_dir, 'tb')
if not os.path.exists(tb_path):
os.mkdir(tb_path)
#tb_path = os.path.join(tb_path, root_dir)
# os.mkdir(tb_path)
tensorboard = TensorBoard(log_dir=tb_path)
callbacks.append(tensorboard)
'''
findlr=False
if findlr:
lr_callback = LRFinder(len(training_generator)*batch_size, batch_size,
1e-05, 1e01,
# validation_data=(X_val, Y_val),
lr_scale='exp', save_dir=join(root_dir,'clr'))
callbacks.append(lr_callback)
oclr=True #CORREGGI STA MERDA OPPURE CANCELLA
if oclr:
lr_manager = OneCycleLR(
max_lr=0.001,
maximum_momentum=0.9,
minimum_momentum=None,
verbose=True
)
callbacks.append(lr_manager)
oclr=True
if oclr:
ocp = ktools.one_cycle.OneCycle(
lr_range=(1e-5, 1e-3),
momentum_range=(0.95, 0.85),
reset_on_train_begin=True,
record_frq=10
)
callbacks.append(ocp)
'''
if clr:
cyclelr = CyclicLR(
base_lr=base_lr,
max_lr=max_lr,
step_size=step_size * len(training_generator),
#mode='exp_range' # uncomment to activate exp mode, instead of traingular
)
callbacks.append(cyclelr)
model.compile(optimizer=optimizer, loss=loss)
if len(val_folders) > 0:
model.fit(x=training_generator,
validation_data=validation_generator,
steps_per_epoch=len(training_generator),
validation_steps=len(validation_generator),
epochs=epochs,
verbose=1,
max_queue_size=10,
use_multiprocessing=True,
workers=workers,
shuffle=False,
callbacks=callbacks)
else:
model.fit(x=training_generator,
steps_per_epoch=len(training_generator),
epochs=epochs,
verbose=1,
max_queue_size=10,
use_multiprocessing=True,
workers=workers,
shuffle=False,
callbacks=callbacks)
# mp.set_start_method('fork')
# save results
train_history = root_dir + '/train-history'
with open(train_history, 'wb') as file_pi:
pickle.dump(history.dic, file_pi)
# post training operations
# training plots
train_plots(train_history, False)
if len(test_dirs) > 0:
if len(val_folders) > 0:
# get the best model on validation
val_loss = history.dic['val_losses']
m = val_loss.index(
min(val_loss)) # get the index with the highest accuracy
model_checkpoints = [
join(root_dir, f) for f in listdir(root_dir)
if (isfile(join(root_dir, f)) and f.startswith(
model_name + '_' + '{:02d}'.format(m + 1)))
]
best = model_checkpoints[0]
print('Best checkpoint: ', best)
else:
# get the best model on validation
acc = history.dic['losses']
m = acc.index(min(acc)) # get the index with the highest accuracy
model_checkpoints = [
join(root_dir, f) for f in listdir(root_dir)
if (isfile(join(root_dir, f)) and f.startswith(
model_name + '_' + '{:02d}'.format(m + 1)))
]
best = model_checkpoints[0]
print('Best checkpoint: ', best)
# test plots & results if test data is provided
if len(test_dirs) > 0:
pkl = tester(test_dirs, best, batch_size, time, feature, workers)
test_plots(pkl, feature)
trainlist = datacfg_options['train']
nsamples = utils.read_train_list(trainlist)
gpus = datacfg_options['gpus']
num_workers = int(datacfg_options['num_workers'])
backupdir = datacfg_options['backup']
if not os.path.exists(backupdir):
os.makedirs(backupdir)
batch_size = int(netcfg_options['batch'])
max_batches = int(netcfg_options['max_batches'])
learning_rate = float(netcfg_options['learning_rate'])
momentum = float(netcfg_options['momentum'])
decay = float(netcfg_options['decay'])
steps = [float(step) for step in netcfg_options['steps'].split(',')]
scales = [float(scale) for scale in netcfg_options['scales'].split(',')]
bg_file_names = utils.get_all_files('./VOCdevkit/VOC2012/JPEGImages')
# Train parameters
max_epochs = 700 # max_batches*batch_size/nsamples+1
use_cuda = True
seed = int(time.time())
eps = 1e-5
save_interval = 10 # epoches
dot_interval = 70 # batches
best_acc = -1
# Test parameters
conf_thresh = 0.05
nms_thresh = 0.4
match_thresh = 0.5
iou_thresh = 0.5
def tester(folders,
model,
batch_size,
time,
feature,
workers,
intensity_cut,
leakage,
gpu_fraction=1,
e_model=None,
test_indexes=None):
if 0 >= gpu_fraction or gpu_fraction > 1:
pass
###################################
#TensorFlow wizardry for GPU dynamic memory allocation
else:
config = tf.ConfigProto()
# Don't pre-allocate memory; allocate as-needed
config.gpu_options.allow_growth = True
# Only allow a fraction of the GPU memory to be allocated
config.gpu_options.per_process_gpu_memory_fraction = gpu_fraction
# Create a session with the above options specified.
keras.backend.tensorflow_backend.set_session(tf.Session(config=config))
###################################
all_files = utils.get_all_files(folders)
print('Building test generator...')
plike = False
if feature == 'direction':
plike = True
test_generator = LSTGenerator(h5files=all_files,
feature=feature,
batch_size=batch_size,
arrival_time=time,
shuffle=False,
intensity=intensity_cut,
leakage2_intensity=leakage,
plike=plike,
load_indexes=test_indexes)
test_idxs = test_generator.get_all_info()
if test_indexes is None:
test_idxs.to_pickle("./test_indexes.pkl")
pd.DataFrame(all_files).to_csv("./test_files.csv")
#if feature == 'gammaness':
# test_gammas = np.unique(test_idxs['class'], return_counts=True)[1][1]
# test_protons = np.unique(test_idxs['class'], return_counts=True)[1][0]
# test_gamma_frac = test_generator.gamma_fraction()
# print('Summary: \nTest gammas: {} \nTest protons: {}, \nGamma fraction: {}'.format(test_gammas, test_protons,test_gamma_frac))
print('Number of test batches: ' + str(len(test_generator)))
print('Loading model...')
mdl = keras.models.load_model(model)
print('Predicting {}...'.format(feature))
predictions = mdl.predict_generator(generator=test_generator,
max_queue_size=10,
workers=workers,
use_multiprocessing=True,
verbose=1)
# adjust length of indexes
print('Retrieving true {}...'.format(feature))
df = pd.DataFrame()
df['energy_true'] = test_generator.get_energies(log=True)
if feature is not 'energy' and e_model is not None:
print('Predicting energy...')
e_mdl = keras.models.load_model(e_model)
e_predictions = e_mdl.predict_generator(generator=test_generator,
max_queue_size=10,
workers=workers,
use_multiprocessing=True,
verbose=1)
df['energy_reco'] = e_predictions
if feature == 'gammaness':
df['class'] = test_generator.get_class()
df['gammaness'] = predictions
elif feature == 'energy':
df['energy_reco'] = predictions
elif feature == 'direction':
df['d_alt_true'] = test_generator.get_AltAz(nom_frame=True)[:, 0]
df['d_az_true'] = test_generator.get_AltAz(nom_frame=True)[:, 1]
df['d_alt_reco'] = predictions[:, 0]
df['d_az_reco'] = predictions[:, 1]
res_file = model + '_' + feature + '_test.pkl'
df.to_pickle(res_file)
print('Results saved in ' + res_file)
#print('Summary: \nTest gammas: {} \nTest protons: {}, \nGamma fraction: {}'.format(test_gammas, test_protons, test_gamma_frac))
return df
# converts video framerates all to specified rate
import utils
import subprocess
import os
all_vids = utils.get_all_files("./data/raw", "mp4")
for v in all_vids:
directory = os.path.split(v)[0]
directory = f'{os.path.split(v)[0]}/video/'
if not os.path.exists(directory):
print(f'making directory: {directory}')
os.makedirs(directory)
filename = os.path.split(v)[-1]
# filename = filename[:-4]
# filename = f"{filename}.wav"
output_file = f'{directory}{filename}'
result = subprocess.Popen(["ffmpeg", "-i", v, "-filter:v", "fps=30", output_file],stdout=subprocess.PIPE)
result.wait()
def build_native_dir(directory, output_dir, cache_dir, abis, std_includes_path, rules: BaseConfig):
executables = {}
for abi in abis:
executable = prepare_compiler_executable(abi)
if executable is None:
print("failed to acquire GCC executable from NDK for abi " + abi)
return CODE_FAILED_NO_GCC
executables[abi] = executable
try:
manifest = get_manifest(directory)
targets = {}
soname = "lib" + manifest["shared"]["name"] + ".so"
for abi in abis:
targets[abi] = os.path.join(output_dir, "so/" + abi + "/" + soname)
except Exception as err:
print("failed to read manifest for directory " + directory + " error: " + str(err))
return CODE_FAILED_INVALID_MANIFEST
keep_sources = rules.get_value("keepSources", fallback=False)
if keep_sources:
# copy everything and clear build files
copy_directory(directory, output_dir, clear_dst=True)
clear_directory(os.path.join(output_dir, "so"))
os.remove(os.path.join(output_dir, soname))
else:
clear_directory(output_dir)
# copy manifest
copy_file(os.path.join(directory, "manifest"), os.path.join(output_dir, "manifest"))
# copy includes
keep_includes = rules.get_value("keepIncludes", fallback=True)
for include_path in manifest["shared"]["include"]:
src_include_path = os.path.join(directory, include_path)
output_include_path = os.path.join(output_dir, include_path)
if keep_includes:
copy_directory(src_include_path, output_include_path, clear_dst=True)
else:
clear_directory(output_include_path)
std_includes = []
for std_includes_dir in os.listdir(std_includes_path):
std_includes.append(os.path.abspath(os.path.join(std_includes_path, std_includes_dir)))
# compile for every abi
overall_result = CODE_OK
for abi in abis:
printed_compilation_title = f"compiling {os.path.basename(directory)} for {abi}"
print("\n")
print(f"{'=' * (48 - len(printed_compilation_title) // 2)} {printed_compilation_title} {'=' * (48 - (1 + len(printed_compilation_title)) // 2)}")
executable = executables[abi]
gcc = [executable, "-std=c++11"]
includes = []
for std_includes_dir in std_includes:
includes.append(f'-I{std_includes_dir}')
dependencies = [f'-L{get_fake_so_dir(abi)}', "-landroid", "-lm", "-llog"]
for link in rules.get_value("link", fallback=[]) + make_config.get_value("make.linkNative", fallback=[]) + ["horizon"]:
add_fake_so(executable, abi, link)
dependencies.append(f'-l{link}')
if "depends" in manifest:
search_dir = os.path.abspath(os.path.join(directory, "..")) # always search for dependencies in current dir
for dependency in manifest["depends"]:
if dependency is not None:
add_fake_so(executable, abi, dependency)
dependencies.append("-l" + dependency)
dependency_dir = search_directory(search_dir, dependency)
if dependency_dir is not None:
try:
for include_dir in get_manifest(dependency_dir)["shared"]["include"]:
includes.append("-I" + os.path.join(dependency_dir, include_dir))
except KeyError:
pass
else:
print(f"ERROR: dependency directory {dependency} is not found, it will be skipped")
# prepare directories
source_files = get_all_files(directory, extensions=(".cpp", ".c"))
preprocessed_dir = os.path.abspath(os.path.join(cache_dir, "preprocessed", abi))
ensure_directory(preprocessed_dir)
object_dir = os.path.abspath(os.path.join(cache_dir, "object", abi))
ensure_directory(object_dir)
# pre-process and compile changes
import filecmp
object_files = []
recompiled_count = 0
for file in source_files:
relative_file = relative_path(directory, file)
sys.stdout.write("preprocessing " + relative_file + " " * 64 + "\r")
object_file = os.path.join(object_dir, relative_file) + ".o"
preprocessed_file = os.path.join(preprocessed_dir, relative_file)
tmp_preprocessed_file = preprocessed_file + ".tmp"
ensure_file_dir(preprocessed_file)
ensure_file_dir(object_file)
object_files.append(object_file)
result = subprocess.call(gcc + ["-E", file, "-o", tmp_preprocessed_file] + includes)
if result == CODE_OK:
if not os.path.isfile(preprocessed_file) or not os.path.isfile(object_file) or \
not filecmp.cmp(preprocessed_file, tmp_preprocessed_file):
if os.path.isfile(preprocessed_file):
os.remove(preprocessed_file)
os.rename(tmp_preprocessed_file, preprocessed_file)
if os.path.isfile(object_file):
os.remove(object_file)
sys.stdout.write("compiling " + relative_file + " " * 64 + "\n")
result = max(result, subprocess.call(gcc + ["-c", preprocessed_file, "-shared", "-o", object_file]))
if result != CODE_OK:
if os.path.isfile(object_file):
os.remove(object_file)
overall_result = result
else:
recompiled_count += 1
else:
if os.path.isfile(object_file):
os.remove(object_file)
overall_result = result
print(" " * 128)
if overall_result != CODE_OK:
print("failed to compile", overall_result)
return overall_result
else:
print(f"recompiled {recompiled_count}/{len(object_files)} files with result {overall_result}")
ensure_file_dir(targets[abi])
command = []
command += gcc
command += object_files
command.append("-shared")
command.append("-Wl,-soname=" + soname)
command.append("-o")
command.append(targets[abi])
command += includes
command += dependencies
print("linking object files...")
result = subprocess.call(command)
if result == CODE_OK:
print("build successful")
else:
print("linker failed with result code", result)
overall_result = result
return overall_result
return overall_result
#!/usr/bin/env python
"""
Renders manual_tests output files
"""
import render
import utils
if __name__ == '__main__':
filenames = utils.get_all_files('manual_tests_output', ['*' + render.INPUT_ARRAY_FORMAT])
filenames.sort()
for filename in filenames:
try:
tags = render.parse_tags(filename)
# Pick the first component only
if render.Y_TAG in tags or render.Z_TAG in tags:
continue
# Pick the first frame only
if render.is_animation(tags) and '0000' not in tags:
continue
if render.POINT2_TAG in tags:
render.render_point2(filename)
elif render.POINT3_TAG in tags:
render.render_point3(filename)
elif render.LINE2_TAG in tags:
render.render_line2(filename)
elif render.LINE3_TAG in tags:
render.render_line3(filename)
elif render.GRID2_TAG in tags:
continue
tweets = pd.DataFrame(tweets_data)
return tweets
def result(total_pos, total_neu, total_neg):
each_live_result = dict(num_pos=total_pos,
num_neu=total_neu,
num_neg=total_neg)
return each_live_result
if __name__ == '__main__':
tweets_data_path = 'tweets_stream/'
out = {}
files = get_all_files(Config.data_path + tweets_data_path,
extension='json')
for file in files:
tweets = load_live_tweets(file)
cleaned_tweets = preprocess(tweets)
total_pos, total_neu, total_neg = vader_sentiment(cleaned_tweets)
time_stamp = cleaned_tweets.iat[
1, 0] # get the time stamp of each live stream file
each_live_result = result(total_pos, total_neu, total_neg)
out.update({time_stamp: each_live_result})
save_to_disk(data=out,
path=Config.reports_path,
filename='live_sentiment_summary.json')
chrome_path_mac = 'open -a /Applications/Google\ Chrome.app %s'
chrome_path_windows = 'C:/Program Files (x86)/Google/Chrome/Application/chrome.exe %s'
if platform.system() == 'Darwin':
web_browser_path = chrome_path_mac
elif platform.system() == 'Windows':
web_browser_path = chrome_path_windows
#Running of Main Code Body with Warning Suppression
with warnings.catch_warnings():
warnings.simplefilter('ignore')
safe_mkdir(output_dir)
safe_mkdir(output_dir_cluster)
#JPEG Colour Converter
print('WITHOUT CLUSTERING: Converting image files to ASCII HTML files now.')
for file_name in get_all_files(input_dir, file_ext='.jpg'):
img_obj = JPEGColourConverter(img_path=file_name, output_dir=output_dir, line_height=line_height, font_size=font_size, h_stretch=h_stretch, symbol=symbol, max_size=max_size, web_browser=web_browser_path, background_colour=background_colour)
img_obj.convert_save_open(open=False, loading_name=img_obj.save_file_name)
print('Conversion completed.')
#JPEG Cluster Colour Converter
print('WITH CLUSTERING: Converting image files to ASCII HTML files now.')
for num_cluster in num_clusters:
for file_name in get_all_files(input_dir, file_ext='.jpg'):
img_obj = JPEGClusterColourConverter(img_path=file_name, output_dir=output_dir_cluster, line_height=line_height, font_size=font_size, h_stretch=h_stretch, symbol=symbol, max_size=max_size, web_browser=web_browser_path, background_colour=background_colour, num_clusters=num_cluster)
img_obj.convert_save_open(open=False, loading_name=img_obj.save_file_name)
print('Conversion completed.')
maindir = sys.argv[1]
output = sys.argv[2]
# sanity checks
if not os.path.isdir(maindir):
print 'ERROR: directory',maindir,'does not exists.'
sys.exit(0)
if os.path.isfile(output):
print 'ERROR: file',output,'exists, delete or provide a new filename.'
sys.exit(0)
# start time
t1 = time.time()
# get all h5 files
allh5 = utils.get_all_files(maindir,ext='.h5')
print 'found',len(allh5),'H5 files.'
# create summary file
HDF5.create_aggregate_file(output,expectedrows=len(allh5),
summaryfile=True)
print 'Summary file created, we start filling it.'
# fill it
h5 = HDF5.open_h5_file_append(output)
HDF5.fill_hdf5_aggregate_file(h5,allh5,summaryfile=True)
h5.close()
# done!
stimelength = str(datetime.timedelta(seconds=time.time()-t1))
print 'Summarized',len(allh5),'files in:',stimelength
# GET DIR/FILE
if not os.path.exists(sys.argv[1]):
print 'file or dir:', sys.argv[1], 'does not exist.'
sys.exit(0)
if os.path.isfile(sys.argv[1]):
if os.path.splitext(sys.argv[1])[1] != '.h5':
print 'we expect a .h5 extension for file:', sys.argv[1]
sys.exit(0)
allh5files = [os.path.abspath(sys.argv[1])]
elif not os.path.isdir(sys.argv[1]):
print sys.argv[
1], "is neither a file nor a directory? confused... a link? c'est klug?"
sys.exit(0)
else:
allh5files = utils.get_all_files(sys.argv[1], ext='.h5')
if len(allh5files) == 0:
print 'no .h5 file found, sorry, check directory you gave us:', sys.argv[
1]
# final sanity checks
for f in allh5files:
assert os.path.splitext(
f
)[1] == '.h5', 'file with wrong extension? should have been caught earlier... file=' + f
nFiles = len(allh5files)
if nFiles > 1000:
print 'you are creating', nFiles, 'new matlab files, hope you have the space and time!'
# let's go!
cnt = 0
for i in chunked:
if type(i) == Tree:
current_chunk.append(" ".join([token for token, pos in i.leaves()]))
if current_chunk:
named_entity = " ".join(current_chunk)
if named_entity not in continuous_chunk:
continuous_chunk.append(named_entity)
current_chunk = []
else:
continue
return continuous_chunk
if __name__ == "__main__":
tweets_data_path = 'tweets_by_country/'
all_files = get_all_files(Config.base_path + tweets_data_path, extension='csv')
out = {}
for country in Config.country_prefix:
df = pd.DataFrame()
news_files = list(filter(lambda x: country in x, all_files))
for file in news_files:
data = pd.read_csv(file, names=Config.colnames, usecols=Config.usecols_list)
data.dropna(axis=0, how='any', inplace=True)
df = df.append(data, ignore_index=True)
text_df = filter_df(Config.keywords, df)
text_list = df_to_str(text_df)
text_str = '. '.join(text_list)
namedEntities = get_continuous_chunks(text_str)
out[country.replace('_', '')] = namedEntities
save_to_disk(data=out,
path=Config.base_path + 'report_named_entities/',
") && "
"cd {0} && "
"git add -A && "
"git commit --amend -q -m Autogenerated && "
"git push -f origin master:gh-pages"
.format(doc_dir, branch),
"deploy-doc",
[doc_rule],
phony=True,
no_format=True)
mk.init_logger()
root = "src"
srcs = list(utils.get_all_files(root))
pp_rules = list(make_preprocess_rules(srcs))
rules, ppfiles = make_copy_header_rules(srcs, [".h", ".hpp"], prefix="calico/")
prepare_rule = mk.alias("prepare", ppfiles)
test_rules = list(make_test_rules(srcs, [".c", ".cpp"], root))
build_rule = mk.alias("build", rules + [prepare_rule])
build_bench_rule = mk.alias("build-bench", [x[2] for x in test_rules
if x[2] is not None])
build_check_rule = mk.alias("build-check", [x[0] for x in test_rules
if x[0] is not None])
maindir = sys.argv[1]
output = sys.argv[2]
# sanity checks
if not os.path.isdir(maindir):
print 'ERROR: directory', maindir, 'does not exists.'
sys.exit(0)
if os.path.isfile(output):
print 'ERROR: file', output, 'exists, delete or provide a new filename.'
sys.exit(0)
# start time
t1 = time.time()
# get all h5 files
allh5 = utils.get_all_files(maindir, ext='.h5')
print 'found', len(allh5), 'H5 files.'
# create summary file
HDF5.create_aggregate_file(output,
expectedrows=len(allh5),
summaryfile=True)
print 'Summary file created, we start filling it.'
# fill it
h5 = HDF5.open_h5_file_append(output)
HDF5.fill_hdf5_aggregate_file(h5, allh5, summaryfile=True)
h5.close()
# done!
stimelength = str(datetime.timedelta(seconds=time.time() - t1))
ES_HOST = sys.argv[3]
start_time = time.time()
logging.basicConfig(format='%(asctime)s %(message)s',
datefmt='%m/%d/%Y %I:%M:%S %p',
filename='./Logs/mterms_data_preparation.log',
level=logging.INFO)
logger = logging.getLogger('LiteratureAnalysis')
logger.info('Started')
db_connection = elasticsearch_connection.ElasticsearchConnection(
ES_HOST, ES_AUTH_USER, ES_AUTH_PASSWORD)
elasticsearch_client = db_connection.get_elasticsearch_client()
dir_path = r"D:/ASU_Part_time/LiteratureAnalysis/TermVectorJstorData/"
extension = "json"
files_to_proceed = utils.get_all_files(
dir_path,
extension)[::-1][len(utils.get_all_files(dir_path, extension)) -
24302:]
for file_number, file in enumerate(files_to_proceed):
completed_files = open("./completed_files_list.csv", 'a')
noncompleted_files = open("./noncompleted_files_list.csv", 'a')
batch_start_time = time.time()
try:
print("Processing of {} file number {} Started ".format(
file.split("\\")[-1], 24302 - file_number - 1))
logging.info(
"Processing of file number {} with filename {} started".format(
24302 - file_number - 1, file))
get_term_vectors(file, elasticsearch_client,
24302 - file_number - 1)
print("Processing of file number {} Completed".format(24302 -
def render_line2(filename):
if is_animation(parse_tags(filename)):
data = []
def get_line_data(filename, frame):
filename = filename.replace('0000', '%04d' % frame)
data_x = np.load(filename)
data_y = np.load(filename.replace(',' + X_TAG, ',' + Y_TAG))
return (data_x, data_y)
def update_lines(frame, line):
x, y = data[frame]
line.set_data(x, y)
return line,
dirname = os.path.dirname(filename)
basename = os.path.basename(filename).replace('0000',
'[0-9][0-9][0-9][0-9]')
seq = utils.get_all_files(dirname, [basename])
seq.sort()
if len(seq) == 0:
return
data_x, data_y = get_line_data(seq[0], 0)
has_bbox = False
for frame in range(len(seq)):
data_x, data_y = get_line_data(seq[frame], frame)
if len(data_x) > 0 and len(data_y) > 0:
if has_bbox:
xmin = min(xmin, data_x.min())
xmax = max(xmax, data_x.max())
ymin = min(ymin, data_y.min())
ymax = max(ymax, data_y.max())
else:
xmin = data_x.min()
xmax = data_x.max()
ymin = data_y.min()
ymax = data_y.max()
has_bbox = True
data.append((data_x, data_y))
fig, ax = plt.subplots()
x, y = get_line_data(filename, 0)
if has_bbox:
xmid = (xmax + xmin) / 2.0
ymid = (ymax + ymin) / 2.0
new_xmin = xmid - 1.25 * (xmid - xmin)
new_xmax = xmid + 1.25 * (xmax - xmid)
new_ymin = ymid - 1.25 * (ymid - ymin)
new_ymax = ymid + 1.25 * (ymax - ymid)
ax.set_xlim(new_xmin, new_xmax)
ax.set_ylim(new_ymin, new_ymax)
ax.set_aspect('equal')
line, = ax.plot(x, y, lw=2, marker='o', markersize=3)
anim = animation.FuncAnimation(fig,
update_lines,
len(seq),
fargs=(line, ),
interval=60,
blit=False)
output_filename = get_output_movie_filename(
filename.replace(',' + X_TAG, ''))
anim.save(output_filename,
fps=60,
bitrate=5000,
writer=video_writer,
extra_args=video_extra_args)
plt.close(fig)
print('Rendered <%s>' % output_filename)
else:
output_filename = get_output_bitmap_filename(
filename.replace(',' + X_TAG, ''))
render_still_line2(filename, filename.replace(',' + X_TAG,
',' + Y_TAG),
output_filename)
import os
import utils
from tag import Tag
from doc_loader import get_data
if __name__ == '__main__':
from sys import argv, stderr
if len(argv) < 2:
print("Usage: {} input_dir output_dir".format(argv[0]), file=stderr)
exit(1)
# Load all python files in the directory (and subdirectories)
output = argv[2]
if not os.path.isdir(output):
os.mkdir(output)
files = utils.get_all_files(argv[1], '.py')
Tag.load_tags('tags.yaml', True)
# Convert all python files into AST objects
classes = []
functions = []
for file in files:
_ast = utils.load_ast(file)
classes.extend(utils.get_classes(_ast))
functions.extend(utils.get_functions(_ast))
# Create the readme, then create all subsequent
data = get_data(classes, functions, output)
readme_template = utils.load_template("../templates/readme.md", True)
md = readme_template.render(data)
_y = []
for xyz in xyz_file:
_x.append(float(xyz.split()[0]))
_y.append(float(xyz.split()[1]))
line.set_data(_x, _y)
xyz_file.close()
return line,
# Simulation Path
path = sys.argv[1]
output_name = sys.argv[2] if len(sys.argv) > 2 else 'result.mp4'
files = utils.get_all_files(path, "*.xyz")
size_files = len(files)
# Set up formatting for the movie files
Writer = animation.writers['ffmpeg']
writer = Writer(fps=25, bitrate=1800)
fig1 = plt.figure()
data = []
l, = plt.plot([], [], 'ro')
plt.xlim(0, 1)
plt.ylim(0, 1)
plt.xlabel('x')
plt.title('Simulation Result')