def compile_by_domain_and_scope(domain, scope):
common.print_verbose("Compiling " + scope)
compile.compile(
"domains/" + domain + "/src/" + scope + "/java",
"target/domains/" + scope + "/" + domain,
dependencies.classpath_for(domain, scope),
)
def add_to_classpath(cp, to_add):
if cp == "":
cp = ":".join(to_add)
else:
cp = cp + ":" + ":".join(to_add)
common.print_verbose("New classpath: " + cp)
return cp
def main(argv):
# Parse arguments
args = parse_args(argv)
set_verbose_level(args.verbose)
print_verbose("Args: %s" % str(args), 1)
# Adapted from
# https://nbviewer.jupyter.org/github/BVLC/caffe/blob/master/examples/00-classification.ipynb
np.set_printoptions(threshold=np.nan)
caffe.set_mode_cpu()
net = caffe.Net(args.proto, args.model, caffe.TEST)
# input preprocessing: 'data' is the name of the input blob == net.inputs[0]
transformer = caffe.io.Transformer({'data': net.blobs['data'].data.shape})
transformer.set_transpose('data', (2, 0, 1))
transformer.set_mean('data', VGG_MEAN_PIXEL) # mean pixel
transformer.set_raw_scale(
'data', 255
) # the reference model operates on images in [0,255] range instead of [0,1]
transformer.set_channel_swap(
'data',
(2, 1,
0)) # the reference model has channels in BGR order instead of RGB
# Read image names
with open(args.list) as f:
allnames = f.read().splitlines()
for sub in xrange(0, len(allnames), CAFFE_BATCH_SIZE):
fnames = allnames[sub:sub + CAFFE_BATCH_SIZE]
# Reshape input data
print net.blobs['data'].data.shape
net.blobs['data'].reshape(len(fnames), *net.blobs['data'].shape[1:])
print net.blobs['data'].data.shape
# Preprocess images
for idx, fname in enumerate(fnames):
fpath = os.path.join(args.input, fname)
print "Processing image %s ..." % fpath
img = transformer.preprocess('data', caffe.io.load_image(fpath))
net.blobs['data'].data[idx] = img
# Extract features
print "Extracting features ..."
out = net.forward()
# Write extracted features
for idx, fname in enumerate(fnames):
path = os.path.join(args.output, os.path.dirname(fname))
if not os.path.exists(path):
os.makedirs(path)
fpath = os.path.join(args.output, fname + ".feat")
print "Writing features to %s ..." % fpath
np.savetxt(fpath, net.blobs['fc7'].data[idx])
print "Done!"
def resize_image(page):
# Global params
global gb_idx_pageid
global gb_idx_img_url
global gb_idx_artist
global gb_idx_realheight
global gb_idx_realwidth
global gb_density
global gb_orig_dir
global gb_dest_dir
# Parse values
pageid = page[gb_idx_pageid]
img_url = page[gb_idx_img_url]
artist = page[gb_idx_artist]
realheight = float(page[gb_idx_realheight])
realwidth = float(page[gb_idx_realwidth])
# Parse paths
orig_path, dest_path = parse_entry_paths(gb_orig_dir, gb_dest_dir, pageid, img_url, artist)
# Parse dimensions
pixelheight, pixelwidth = parse_entry_sizes(gb_density, realheight, realwidth)
# Call convert to resize image
convert_resize(orig_path, dest_path, pixelheight, pixelwidth)
# Call convert to update density metadata
convert_density(dest_path, realheight, realwidth)
print_verbose('Done processing pageid %s' % pageid, 0)
def get_vm_list(self, filter_params={}):
p_instance = filter_params.get("instance", "")
self.header_keys = ['region', 'name', 'type']
self.header_titles = ['Region', 'Name', 'Instance Type']
vms = []
for avm in self.client.virtual_machines.list_all():
vm = AzureVM(avm)
if p_instance > "":
if p_instance != vm.name:
continue
else:
dm = self.client.virtual_machines.get(
vm.resource_group, vm.name, expand="instanceView")
vm.set_extra_info(dm.__dict__)
for n in dm.network_profile.network_interfaces:
name = " ".join(n.id.split('/')[-1:])
sub = "".join(n.id.split('/')[4])
ips = self.networkClient.network_interfaces.get(
sub, name)
vm.set_extra_info(ips.__dict__)
if vm.public_ip_address is None:
ip_reference = ips.ip_configurations[
0].public_ip_address.id.split('/')
public_ip = self.networkClient.public_ip_addresses.get(
ip_reference[4], ip_reference[8])
vm.public_ip_address = public_ip.ip_address
if self.verbose:
print_verbose(vm.__dict__, p_type="vm")
vms.append(vm.__dict__)
return vms
def download_image(img_path, img_url):
# Fetch URL
url = urllib2.urlopen(img_url)
meta = url.info()
file_size = int(meta.getheaders("Content-Length")[0])
print_verbose("Downloading image %s (%s)" % (url.geturl(),
size(file_size, system=alternative)), 0)
# Set progress bar
widgets = ['Progress: ', Percentage(), ' ', Bar(),
' ', AdaptiveETA(), ' ', AdaptiveTransferSpeed()]
pbar = ProgressBar(widgets=widgets, maxval=file_size).start()
# Download
f = open(img_path, 'wb')
file_size_dl = 0
block_sz = 1024 * 8
while True:
buff = url.read(block_sz)
if not buff:
break
file_size_dl += len(buff)
f.write(buff)
pbar.update(file_size_dl)
# Done
f.close()
pbar.finish()
return url.getcode()
def __init__(self, dependencies_file_contents):
self.dependencies_file_contents = dependencies_file_contents
self.dependencies = dict()
dependencies_yaml = yaml.load(self.dependencies_file_contents)
if not dependencies_yaml:
common.print_verbose("No dependencies found")
return
for stage in dependencies_yaml.items():
stage_name_from_yaml = stage[0]
# No stages defined
if not stage_name_from_yaml in self.dependencies:
self.dependencies[stage_name_from_yaml] = list()
# Stage is defined but not dependencies listed for the stage
stage_dependencies_from_yaml = stage[1] if stage[1] else dict()
for dependency in stage_dependencies_from_yaml.items():
dep_name = dependency[0]
dep_version = dependency[1]["version"]
dep_installer = dependency[1]["installer"]
dep_details = dict(dependency[1])
del dep_details["version"]
del dep_details["installer"]
self.dependencies[stage_name_from_yaml].append(\
Dependency(name=dependency[0], version=dep_version, installer=dep_installer, details=dep_details))
return
def run(self):
common.print_verbose("Running " + self.name + " action")
self.deps = self.load_dependencies(self.dependencies_file_location)
for dep in self.deps.dependencies_for("default"):
dep.install()
return 0, ""
def convert_resize(orig_path, dest_path, pixelheight, pixelwidth):
# Prepare command
cmd = ["convert", orig_path, "-resize",
"%dx%d^" % (pixelwidth, pixelheight), dest_path]
if get_verbose_level() >= 5:
cmd.insert(1, "-verbose")
# Run
print_verbose("Running command: " + list2cmdline(cmd), 3)
return check_call(cmd)
def main(argv):
# Parse arguments
args = parse_args(argv)
set_verbose_level(args.verbose)
print_verbose("Args: %s" % str(args), 1)
# Resize images
resize_from_csv(args.csv, args.original, args.resized, args.density)
def main(argv):
# Parse arguments
args = parse_args(argv)
set_verbose_level(args.verbose)
print_verbose("Args: %s" % str(args), 1)
# Download images
download_from_csv(args.csv, args.directory)
def main(argv):
# Parse arguments
args = parse_args(argv)
set_verbose_level(args.verbose)
print_verbose("Args: %s" % str(args), 1)
# Crawl URL
crawl(args.url)
def run(self):
common.print_verbose("Running " + self.name + " action")
with open(self.actions_file_location) as f:
actions_file_contents = f.read()
for stage in yaml.load(actions_file_contents).items():
for an_action in stage[1]:
common.stop_if_failed(*self.action_for(an_action).run())
f.closed
return 0, ""
def run(self):
common.print_verbose("Running " + self.name + " action")
files_to_copy = glob.glob(MetaModel(common.meta_model()).template_for_action(self.name) + "/*", recursive=False)
command_to_run = ['/bin/cp', "-R", *files_to_copy, '.']
common.run_command(command_to_run)
common.print_raw("Initialized new Java 9 application.")
return 0, ""
def parse_class(filename):
if filename.startswith(VG_PREFIX):
cl = VG_CLASS
elif filename.startswith(NVG_PREFIX):
cl = NVG_CLASS
else:
raise ValueError('Invalid class prefix.'
'Valid prefixes are "%s" and "%s"' % (VG_PREFIX, NVG_PREFIX))
print_verbose('File %s class: %d' % (filename, cl), 4)
return cl
def agg_pred_dist_far(pred, classes):
print_verbose('Aggregating by using the farthest point class ...', 0)
arr_pos = pred[pred >= 0]
max_pos = np.max(arr_pos) if (arr_pos.size > 0) else 0
arr_neg = pred[pred <= 0]
max_neg = np.abs(np.min(arr_neg)) if (arr_neg.size > 0) else 0
cl = classes[1] if (max_pos > max_neg) else classes[0]
return cl
def agg_pred_dist_mediangroup(pred, classes):
print_verbose('Aggregating by comparing distance groups medians ...', 0)
arr_pos = pred[pred >= 0]
med_pos = np.median(arr_pos) if (arr_pos.size > 0) else 0
arr_neg = pred[pred <= 0]
med_neg = np.abs(np.median(arr_neg)) if (arr_neg.size > 0) else 0
cl = classes[1] if (med_pos > med_neg) else classes[0]
return cl
def main(argv):
# Parse arguments
args = parse_args(argv)
set_verbose_level(args.verbose)
set_n_cores(args.cores)
print_verbose("Args: %s" % str(args), 1)
# Extract patches
patch_extract(args.image, args.window, args.step, args.dir)
def setup(topo, defaults):
common.null_to_string(topo.addressing)
addrs = setup_pools(defaults.addressing + topo.addressing, defaults)
common.print_verbose("Addressing\n=================")
common.print_verbose(addrs.to_yaml())
validate_pools(addrs)
common.exit_on_error()
topo.pools = create_pool_generators(addrs)
topo.addressing = addrs
common.exit_on_error()
def setup(topo, defaults={}):
addrs = topo.get('addressing', {})
addrs = common.merge_defaults(addrs, defaults.get('addressing'))
addrs = setup_pools(addrs, defaults)
topo['addressing'] = addrs
common.print_verbose("Addressing\n=================")
common.print_verbose(yaml.dump(addrs))
validate_pools(addrs)
common.exit_on_error()
topo['pools'] = create_pool_generators(addrs)
common.exit_on_error()
def run(self):
common.print_verbose("Running " + self.name + " action")
cp_string = ""
if self.classpath != "":
cp_string = " -cp " + self.classpath + " "
common.run_command_in_shell('rm -rf ' + self.target_dir)
common.run_command_in_shell('mkdir -p ' + self.target_dir)
common.run_command_in_shell('find ' + self.source_dir + \
' -type f -name "*.java" -print | xargs javac ' + \
cp_string + \
' -d ' + self.target_dir + ' -sourcepath ' + self.source_dir)
return 0, ""
def run(self):
common.print_verbose("Running " + self.name + " action")
exit_code = 0
for test_dir in glob.iglob('**/test', recursive=True):
original_working_directory = os.getcwd()
run_directory = os.path.join(original_working_directory, str(test_dir))
common.print_info("Running tests in " + str(run_directory))
common.print_verbose("Changing directory to " + str(run_directory))
os.chdir(run_directory)
tests = []
for filename in glob.iglob('**/*.py', recursive=True):
tests.append(filename)
command = ['/usr/local/bin/python3', '-m', 'unittest']
command.extend(tests)
subprocess_exit_code, output = common.run_command(command)
if subprocess_exit_code != common.SUCCESS:
exit_code = common.FAILED
common.print_verbose(output)
common.continue_if_failed(subprocess_exit_code, output)
common.print_verbose("Changing directory to " + str(original_working_directory))
os.chdir(original_working_directory)
return exit_code, ""
def calc_dist(clf_model, data):
# Read model
with open(clf_model, "rb") as f:
model = pickle.load(f)
print_verbose(
"Model [%0.2f%%]: %s" %
(model.best_score_ * 100, str(model.best_estimator_)), 4)
# Calculate distances
dist = model.decision_function(data)
dist = dist.reshape(-1, 1)
return dist
def get_vm_list(self, filter_params={}):
self.header_keys = [
'region', 'name', 'type', 'public_ips', 'private_ips', 'state'
]
self.header_titles = [
'Region', 'Name', 'Instance Type', 'Public Ips', 'Private Ips',
'state'
]
p_region = filter_params.get("region", "")
p_instance = filter_params.get("instance", "")
info = False
if p_instance:
info = True
ec2_regions = []
available_ec2_regions = boto3.session.Session().get_available_regions(
self.instance_type)
if p_region > "":
if p_region in available_ec2_regions:
ec2_regions = [p_region]
else:
msg = str(p_region) + " is not a valid region for ec2."
exit_message(msg, 1, self.isJson)
else:
ec2_regions = available_ec2_regions
ec2_hosts = []
for region in ec2_regions:
try:
msg = "Searching " + region + "..."
message(msg, "info", self.isJson)
ec2 = get_client('ec2', region=region)
hosts = ec2.describe_instances()
for host in hosts['Reservations']:
for h in host['Instances']:
if p_instance > "":
if p_instance != h['InstanceId']:
continue
vm = AwsVM(h)
vm.region = region
if p_instance:
vm.set_extra_info(h)
if self.verbose:
print_verbose(vm.__dict__)
ec2_hosts.append(vm.__dict__)
except Exception as e:
pass
return ec2_hosts
def run_tests_for(domain, scope):
common.print_info_no_eol("Running " + scope + " for " + domain + "...")
classpath = dependencies.classpath_for(domain, scope)
test_classes_as_string = test_classes_for(domain, scope)
if test_classes_as_string.strip() != "":
run_tests_command = "java -cp " + classpath + " org.junit.runner.JUnitCore " + test_classes_as_string
common.print_verbose("Running tests with:")
common.print_verbose(run_tests_command)
(exit_code, output) = common.run_command(run_tests_command)
if exit_code == 0:
common.print_info(" PASSED.")
else:
common.print_info(" FAILED.")
else:
common.print_info(" No tests found.")
def direct_dependencies(domain, scope):
global dependency_versions
dependency_file = "domains/" + domain + "/src/" + scope + "/java/dependencies.csv"
dependencies_to_return = {}
if os.path.isfile(dependency_file):
f = open(dependency_file, "r")
deps = f.read().split("\n")
deps.remove("groupId,artifactId")
for dep in deps:
common.print_verbose("Found dependency: " + dep + " with version " + dependency_versions[dep])
dependencies_to_return[dep] = dependency_versions[dep]
common.print_verbose(dependencies_to_return)
return dependencies_to_return
def main(log_level="info", meta_model_name="power-daps/python3", actions_to_run=["default"]):
common.set_log_level(log_level)
meta_model = MetaModel(meta_model_name)
common.set_meta_model(meta_model_name)
valid_actions = meta_model.actions()
common.print_verbose('Actions to run ' + str(actions_to_run))
common.print_verbose('Valid actions ' + str([va.name for va in valid_actions]))
for action_to_run in actions_to_run:
if action_to_run not in [va.name for va in valid_actions]:
common.print_error("Action '" + action_to_run + "' not found.")
continue
for valid_action in valid_actions:
if valid_action.name == action_to_run:
common.stop_if_failed(*valid_action.run())
def get_rds_list(self, filter_params={}):
self.header_keys = [
'region', 'instance', 'status', 'db_class', 'engine_version'
]
self.header_titles = [
'Region', 'Instance', 'Status', 'Class', 'Version'
]
p_region = filter_params.get("region", "")
p_instance = filter_params.get("instance", "")
rds_regions = []
available_rds_regions = boto3.session.Session().get_available_regions(
"rds")
if p_region > "":
if p_region in available_rds_regions:
rds_regions = [p_region]
else:
msg = str(p_region) + " is not a valid region for rds."
exit_message(msg, 1, self.isJson)
else:
rds_regions = available_rds_regions
# get all of the postgres db instances
pg_list = []
for region in rds_regions:
msg = "Searching " + region + "..."
message(msg, "info", self.isJson)
rds = get_client('rds', region=region)
dbs = rds.describe_db_instances()
for db in dbs['DBInstances']:
if db['Engine'] in ["postgres"]:
if p_instance > "":
if p_instance != db['DBInstanceIdentifier']:
continue
rds = AwsRDS(db)
rds.region = region
if p_instance:
extras_args = {}
extras_args['ec2'] = get_client('ec2', region=region)
rds.set_extra_info(db, extras_args)
if self.verbose:
print_verbose(rds.__dict__, p_type="pg")
pg_list.append(rds.__dict__)
return pg_list
def main(argv):
# Parse arguments
args = parse_args(argv)
set_verbose_level(args.verbose)
print_verbose("Args: %s" % str(args), 1)
# Crawl URL
result = crawl(args.url)
# Extract data
raw_data = extract_data(result)
# Sort (optional)
raw_data.sort()
# Save to CSV
gen_csv(args.csv, raw_data)
def convert_density(filepath, realheight, realwidth):
# Get image dimensions in pixels
pixelheight, pixelwidth = identify_size(filepath)
# Calculate densities
# Densities are calculated in CM due to precision errors in ImageMagick
dens_height = 1.0 * pixelheight / realheight / 2.54
dens_width = 1.0 * pixelwidth / realwidth / 2.54
# Prepare command
cmd = ["convert", "-units", "PixelsPerCentimeter", filepath,
"-density", "%.5fx%.5f" % (dens_width, dens_height),
filepath]
if get_verbose_level() >= 5:
cmd.insert(1, "-verbose")
# Run
print_verbose("Running command: " + list2cmdline(cmd), 3)
return check_call(cmd)
def get_db_list(self, filter_params={}):
p_instance = filter_params.get("instance", "")
self.header_keys = [
'region', 'instance', 'status', 'db_class', 'engine_version'
]
self.header_titles = [
'Region', 'Instance', 'Status', 'Class', 'Version'
]
db_list = []
for pg in self.client.servers.list():
db = AzurePG(pg)
if p_instance > "":
if p_instance != db.instance:
continue
if self.verbose:
db.set_extra_info(pg)
print_verbose(db.__dict__, p_type="pg")
db_list.append(db.__dict__)
return db_list
def gen_data(dirname, gtruth=True):
files = list_files(dirname)
full_paths = map(lambda x: os.path.join(dirname, x), files)
print_verbose('Dir %s full file path: %s' % (dirname, str(full_paths)), 4)
data = apply_multicore_function(read_data, full_paths)
labels = apply_multicore_function(parse_label, files)
if gtruth:
classes = apply_multicore_function(parse_class, files)
data = np.asarray(data, dtype=np.float)
labels = np.asarray(labels, dtype=np.str)
if gtruth:
classes = np.asarray(classes, dtype=np.uint8)
if gtruth:
return data, labels, classes
else:
return data, labels
def load_actions_from_dir(self, dir):
common.print_verbose("Looking for actions in " + dir)
if os.path.isdir(dir) is not True:
common.exit_with_error_message("Meta-model '" + self.name() + "' not found in '" + dir + "'")
elif os.path.isdir(os.path.join(dir, "actions")) is not True:
common.exit_with_error_message("Meta-model '" + self.name() + "' found but no actions found")
elif len(self.actions_found_in(dir + "/actions")) == 0:
common.exit_with_error_message("No actions found in '" + dir + "/actions'")
if dir not in sys.path:
sys.path.insert(0, dir)
actions = []
#for action in ["default", "deps", "unit_test", "package", "run"]:
for action in self.actions_found_in(dir + "/actions"):
action_module = importlib.import_module("actions." + action + "_action")
actions.append(action_module.action())
return actions
def delete_many_parallel(groups):
common.print_info(
"Deleting ResourceGroup in parallel started. May take couple of minutes"
)
# Start processes and redirect output to tmp files
processes = []
for group in groups:
group_name = group["name"]
p = common.process_start(f"az group delete --name {group_name} --yes")
common.print_info(
f"Deleting ResourceGroup {group_name} started. PID={p.pid}")
processes.append(p)
# Wait for processes and print their outputs
common.print_info("Waiting to finish...")
for p in processes:
p.wait()
common.print_verbose(f"{p.pid} finished. Return status={p.returncode}")
for line in p.stdout.readlines():
common.print_info(line)
for line in p.stderr.readlines():
common.print_error(line)
common.print_info("Deleting ResourceGroups (parallel) finished")
def resize_from_csv(csvfile, orig_dir, dest_dir, density):
# Global params
global gb_idx_pageid
global gb_idx_img_url
global gb_idx_artist
global gb_idx_realheight
global gb_idx_realwidth
global gb_density
global gb_orig_dir
global gb_dest_dir
# Define writer
reader = csv.reader(csvfile, quoting=csv.QUOTE_ALL, strict=True)
# Field names
field_names = reader.next()
# Indices
gb_idx_pageid = field_names.index('PageID')
gb_idx_img_url = field_names.index('ImageURL')
gb_idx_artist = field_names.index('Artist')
gb_idx_realheight = field_names.index('RealHeightInches')
gb_idx_realwidth = field_names.index('RealWidthInches')
# Prepare multicore execution
gb_orig_dir = orig_dir
gb_dest_dir = dest_dir
gb_density = density
pool = Pool()
pool.map(resize_image, reader)
pool.close()
pool.join()
print_verbose('Completed!', 0)
def crawl(url_param):
# Fix eventual full URL
url_param = unquote_plus(basename(url_param))
# Generate query
params = {
'action' : 'query',
'prop' : 'imageinfo|revisions',
'iiprop' : 'url|sha1|size',
'rvprop' : 'content',
'rawcontinue' : '' }
url_type = get_url_type(url_param)
if url_type == 'category':
params['generator'] = 'categorymembers'
params['gcmtitle'] = url_param
params['gcmlimit'] = 'max'
elif url_type == 'file':
params['titles'] = url_param
else:
params['generator'] = 'images'
params['titles'] = url_param
params['gimlimit'] = 'max'
# Call API
site = wiki.Wiki(API_URL)
request = api.APIRequest(site, params)
print_verbose("Site: %s" % str(site), 2)
print_verbose("Query: ", 2)
pprint_verbose(params, 2)
result = request.query(querycontinue=True)
print_verbose("Result: ", 4)
pprint_verbose(result, 4)
# Check result
if 'error' in result:
raise Error(result['error'])
if 'warnings' in result:
sys.stderr.write(result['warnings'])
return None
if '-1' in result['query']['pages']:
sys.stderr.write(result['query']['pages']['-1'])
return None
return result['query']['pages']
def fetch(self):
common.print_info_no_eol( " " + self.group_id + "," + self.artifact_id + ": " + self.version + "... ")
common.print_verbose('')
common.print_verbose('URL: ' + self.remotelocation("jar"))
common.print_verbose('Local: ' + self.locallocation("jar"))
if not self.islatest():
self.forcefetch("jar")
self.forcefetch("pom")
self.forcefetch("pom.asc")
self.forcefetch("jar.asc")
common.print_info("downloaded")
else:
common.print_info("exists")
def extract_data(result):
img_list = []
# For each page
for page in result.values():
# Get ID
page_id = page['pageid']
print_verbose("Extracting info from %s ..." % page_id, 1)
try:
# Image info
img_info = page['imageinfo'][0]
img_desc_url = img_info['descriptionurl']
img_url = img_info['url']
img_sha1 = img_info['sha1']
img_height = img_info['height']
img_width = img_info['width']
# Content
img_content = page['revisions'][0]['*']
(paint_id, artist, dim) = extract_image_content(img_content)
print_verbose("URL is %s" % img_desc_url, 1)
# Internal object
img_page = ImagePage(page_id, img_desc_url, img_url, img_sha1,
img_height, img_width,
paint_id, artist, dim)
print_verbose(img_page, 3)
img_list.append(img_page)
except Exception as e:
sys.stderr.write("Error processing PageID %s\n" % page_id)
sys.stderr.write("-- %s\n" % e)
return img_list
def generate_model(data, classes, args):
# Define the parameters
tuned_parameters = {'C': C_RANGE, 'class_weight': CLASS_WEIGHTS}
# Define the classifier
clf = linear_model.LogisticRegression(max_iter=SCORE_MAX_ITER,
n_jobs=args.cores)
print_verbose("Classifier: %s" % str(clf), 5)
print_verbose("Parameters: %s" % str(tuned_parameters), 5)
# Generate the K-fold development
skf = cross_validation.StratifiedKFold(classes,
n_folds=SCORE_K_FOLD,
shuffle=True)
print_verbose("KFold: %s" % str(skf), 5)
gscv = grid_search.GridSearchCV(clf,
tuned_parameters,
cv=skf,
scoring='mean_squared_error',
n_jobs=1,
verbose=get_verbose_level())
# Search
print_verbose("GridSearch: %s" % str(gscv), 5)
gscv.fit(data, classes)
# Print scores
print_verbose("GridSearch scores:", 5)
for params, mean_score, scores in gscv.grid_scores_:
print_verbose(
"%0.6f (+/-%0.06f) for %r" %
(mean_score, scores.std() / 2, params), 5)
# Print best score
print_verbose("GridSearch best score:", 0)
print_verbose("%0.6f for %r" % (gscv.best_score_, gscv.best_params_), 0)
return gscv
def main(argv):
# Parse arguments
args = parse_args(argv)
set_verbose_level(args.verbose)
set_n_cores(args.cores)
print_verbose("Args: %s" % str(args), 1)
# Prepare data
data, labels, classes = gen_data(args.dir)
print_verbose('Data: %s' % str(data), 5)
print_verbose('Labels: %s' % str(labels), 4)
print_verbose('Classes: %s' % str(classes), 4)
print_verbose('Data shape: %s' % str(data.shape), 2)
print_verbose('Labels shape: %s' % str(labels.shape), 2)
print_verbose('Classes shape: %s' % str(classes.shape), 2)
print_verbose('Data bytes: %s' % str(data.nbytes), 2)
# Calculate distances
dist = calc_dist(args.model, data)
# Generate score model
model = generate_model(dist, classes, args)
print_verbose('Model: %s' % str(model), 0)
# Export
print_verbose('Saving model to %s' % args.score, 0)
with open(args.score, "wb") as f:
pickle.dump(model, f)
print_verbose('Done!', 0)
def eval_perf(classification):
y_true = []
y_pred = []
for (key, value) in classification.iteritems():
y_true.extend([parse_class(key)])
y_pred.extend([value])
print_verbose("Classification pair: %s" % str((key, value)), 4)
print_verbose("True classes: %s" % str(y_true), 5)
print_verbose("Predicted classes: %s" % str(y_pred), 5)
# Print results
print_verbose("True classes: %s" % str(y_true), 2)
print_verbose("Predicted classes: %s" % str(y_pred), 2)
# Print metrics
print_verbose("Confusion Matrix:", 0)
print_verbose(metrics.confusion_matrix(y_true, y_pred), 0)
print_verbose("Classification Report:", 0)
print_verbose(metrics.classification_report(y_true, y_pred), 0)
def main(argv):
# Parse arguments
args = parse_args(argv)
set_verbose_level(args.verbose)
set_n_cores(args.cores)
print_verbose("Args: %s" % str(args), 1)
# Some tests
data, labels = gen_data(args.dir, False)
print_verbose('Data: %s' % str(data), 5)
print_verbose('Labels: %s' % str(labels), 4)
print_verbose('Data shape: %s' % str(data.shape), 2)
print_verbose('Labels shape: %s' % str(labels.shape), 2)
classification = classify(data, labels, args)
print_verbose('Final classification: %s' % str(classification), 0)
# Evaluate performance
if args.gtruth:
eval_perf(classification)
def classify(data, labels, args):
classification = {}
# Read model
with open(args.model, "rb") as f:
model = pickle.load(f)
print_verbose(
"Model [%0.2f%%]: %s" %
(model.best_score_ * 100, str(model.best_estimator_)), 4)
# Classify each label
lolo = cross_validation.LeaveOneLabelOut(labels)
print_verbose("LeaveOneOut: %s" % str(lolo), 5)
for train_index, test_index in lolo:
print_verbose("Test index: %s" % str(test_index), 5)
print_verbose("Classifying label: %s" % str(labels[test_index[0]]), 4)
# Classify
if args.aggregation == 'mode':
pred = model.predict(data[test_index])
else:
pred = model.decision_function(data[test_index])
print_verbose("Patch prediction: %s" % str(pred), 4)
# Aggregate
if args.aggregation == 'mode':
res = agg_pred_mode(pred)
elif args.aggregation == 'sum':
res = agg_pred_dist_sumall(pred, model.best_estimator_.classes_)
elif args.aggregation == 'far':
res = agg_pred_dist_far(pred, model.best_estimator_.classes_)
elif args.aggregation == 'mean':
res = agg_pred_dist_meangroup(pred, model.best_estimator_.classes_)
elif args.aggregation == 'median':
res = agg_pred_dist_mediangroup(pred,
model.best_estimator_.classes_)
print_verbose("Aggregate result: %s" % str(res), 4)
# Append to final result
classification[labels[test_index[0]]] = res
print_verbose("Classification: %s" % str(classification), 5)
return classification
def run(self):
common.print_verbose("Running " + self.name + " action")
common.stop_if_failed(*common.run_command(["/bin/rm", "-rf", "dist/dap"]))
return common.run_command([self.pyinstaller(),
"--noconfirm", "--log-level=WARN",
common.power_daps_dir() + "dap.spec"])
def run(self):
common.print_verbose("Running " + self.name + " action")
return 0, ""
def run(self):
common.print_verbose("Running " + self.name + " action")
common.print_raw("blueee!!!")
return 0, ""
