def main(mode: str):
"""
Main program. Either fetches new articles from RSS feeds or analyzes them using Textrazor.
:param mode: One of 'fetch' or 'analyze'.
:return: Nothing, only prints messages regarding success or failure.
"""
# time-related
now = datetime.now()
# configuration, database connection
conf = Config()
db = DatabaseClient(conf.database.host, conf.database.database,
conf.database.user, conf.database.password,
conf.database.port)
# fetch / analyze articles
if mode == "fetch":
fetch(conf, db, now.isoformat())
elif mode == "analyze":
analyze(conf, db, now)
else:
raise ValueError("--mode must be one of 'fetch', 'analyze'.")
# commit any changes to database
db.connection.commit()
def test_analyze(self):
self.assertEqual(analyze([]),
{"avg": 0.0, "n": 0, "variance": 0.0})
self.assertEqual(analyze([3.0]),
{"avg": 3.0, "n": 1, "variance": 0.0})
self.assertEqual(analyze([1.0, 2.0, 3.0, 4.0]),
{"avg": 2.5, "n": 4, "variance": 5.0/3.0})
self.assertEqual(analyze([1.0, 2.0, 3.0, 4.0], linear_weights),
{"avg": 2.0, "n": 4, "variance": 2.0})
def main():
"""
main executable
check times, put into intervals
"""
# load parsed data
data, driver_hash = getData()
# analyze data
analyze(data, driver_hash)
def processApps(args, gpapi, apps):
""" Download and analyze apps on the Google Play Store.
Arguments:
args -- the command line arguments object
gpapi -- the Google Play API object
apps -- dictionary of apps and their meta data
"""
createAppFolder(args)
i = 0
j = 0
for app in apps:
if shouldProcess(apps[app]):
j += 1
print "found {:,} apps to process".format(j)
pos = getRestorePoint(args)
for app,meta in apps.iteritems():
# we only care about apps which require INTERNET permission, we haven't checked yet, and are free
if not shouldProcess(meta):
continue
# skip until at the position where we should resume
i += 1
if i < pos:
continue
# create restore point
if i % args.restore_freq == 0 and i > 0 and args.restore_freq > 0:
createRestorePoint(args, apps, i)
# print progress
sys.stdout.write("\rprocessing apps... %6.2f%% %10s: %s\033[K " % (100.0 * i / j, "app", app))
sys.stdout.flush()
try:
fname = args.app_dir + app + ".apk"
if download(gpapi, fname, app, meta['version'], meta['offer']):
analyze(apps, fname, app)
os.remove(fname)
except:
None
sys.stdout.write("\rdone processing apps\033[K\n")
sys.stdout.flush()
# clean up
print "saving to cache"
clearRestorePoint(args, apps)
deleteAppFolder(args)
def test_analyze(self):
self.assertEqual(analyze([]), {"avg": 0.0, "n": 0, "variance": 0.0})
self.assertEqual(analyze([3.0]), {"avg": 3.0, "n": 1, "variance": 0.0})
self.assertEqual(analyze([1.0, 2.0, 3.0, 4.0]), {
"avg": 2.5,
"n": 4,
"variance": 5.0 / 3.0
})
self.assertEqual(analyze([1.0, 2.0, 3.0, 4.0], linear_weights), {
"avg": 2.0,
"n": 4,
"variance": 2.0
})
def get(self):
a = analyze()
header = """
<html>
<head>
<title>Green Tweets</Title>
"""
self.write(header)
for tweet in tweetstore:
screen_name = ""
text = ""
if a.cleanstr(tweet['text']) == '':
#print "Skipping: " + tweet['text']
pass
else:
if tweet.has_key('text'):
text = tweet['text'].encode('ascii', 'ignore')
text = convertLinks(text)
if tweet.has_key('screen_name'):
screen_name = "@" + tweet['screen_name'].encode(
'ascii', 'ignore') + ": "
elif tweet.has_key('user'):
if tweet['user'].has_key('screen_name'):
screen_name = "@" + tweet['user'][
'screen_name'].encode('ascii', 'ignore') + ": "
if a.finduser(screen_name) or (
a.findword(text)) or a.findterm(text):
self.write(tweet)
self.write("\n")
finhdr = """
</head>
<body style="background:lightgray">
"""
self.write("</body></html>")
def get(self):
a = analyze()
header = """
<html>
<head>
<title>Twitter Client</Title>
<meta http-equiv="refresh" content="60" />
</head>
<body style="background:lightgray">
"""
self.write(header)
for tweet in rtstore:
if tweet.has_key('text'):
text = tweet['text'].encode('ascii', 'ignore')
text = convertLinks(text)
if tweet.has_key('screen_name'):
screen_name = "@" + tweet['screen_name'].encode(
'ascii', 'ignore') + ": "
elif tweet.has_key('user'):
if tweet['user'].has_key('screen_name'):
screen_name = "@" + tweet['user']['screen_name'].encode(
'ascii', 'ignore') + ": "
self.write("<p>")
self.write(screen_name)
self.write(text)
self.write("</p>")
self.write("</body></html>")
def __init__(self):
QMainWindow.__init__(self)
self.setAttribute(QtCore.Qt.WA_DeleteOnClose)
self.setWindowTitle("application main window")
self.file_menu = QMenu('&File', self)
self.file_menu.addAction('&Quit', self.fileQuit,
QtCore.Qt.CTRL + QtCore.Qt.Key_Q)
self.menuBar().addMenu(self.file_menu)
self.help_menu = QMenu('&Help', self)
self.menuBar().addSeparator()
self.menuBar().addMenu(self.help_menu)
self.help_menu.addAction('&About', self.about)
self.main_widget = QWidget(self)
file = "D:/Users/User/Documents/GitHub/gg/gggg.pcap"
df = analyze(file, self)
ax = create_plot(df, )
l = QVBoxLayout(self.main_widget)
sc = MyStaticMplCanvas(ax,
self.main_widget,
width=5,
height=4,
dpi=100)
l.addWidget(sc)
self.main_widget.setFocus()
self.setCentralWidget(self.main_widget)
self.statusBar().showMessage("All hail matplotlib!", 2000)
def process(packets):
packets_to_analyze = []
packets_to_analyze_index = []
packet_index = 0
for packet in packets:
if Raw in packet:
packets_to_analyze.append(packet[Raw].load)
packets_to_analyze_index.append(packet_index)
packet_index += 1
strings_to_analyze = [''] * len(packets_to_analyze)
index = -1
for load in packets_to_analyze:
first_char = ord(load[0])
if first_char < 20 and load.find("json") != -1:
index += 1
strings_to_analyze[index] = load
else:
string = strings_to_analyze[index]
string += load
strings_to_analyze[index] = string
token = -2
for load in strings_to_analyze:
tmp_token = analyze(load)
if tmp_token > 0:
token = tmp_token
if token > 0:
for packet in packets:
if Raw in packet:
str_token = hex(token)[2:len(hex(token))-1]
token += 1
str_token_plus = hex(token)[2:len(hex(token))-1]
packet[Raw].load = packet[Raw].load.replace(str_token, str_token_plus)
def get(self):
a = analyze()
header = """
<html>
<head>
<title>Green Tweets</Title>
</head>
<body style="background:lightgray">
"""
self.write(header)
for tweet in reversed(greenstore):
screen_name = ""
text = ""
if tweet.has_key('text'):
text = tweet['text'].encode('ascii', 'ignore')
text = convertLinks(text)
if tweet.has_key('screen_name'):
screen_name = "@" + tweet['screen_name'].encode('ascii', 'ignore') + ": "
elif tweet.has_key('user'):
if tweet['user'].has_key('screen_name'):
screen_name = "@" + tweet['user']['screen_name'].encode('ascii', 'ignore') + ": "
img_url = ""
if tweet['user'].has_key('profile_image_url'):
img_url = tweet['user']['profile_image_url']
self.write("<p style='background:white; color: green'>")
if img_url != "":
self.write("<img src='" + img_url + "' style='float:left'>")
self.write("<a href='http://twitter.com/" + screen_name[1:-2] + "' target='_blank'>" + screen_name + "</a>")
self.write("<br/>")
self.write(text)
if tweet.has_key('created_at'):
self.write("<br/>" + tweet['created_at'].encode('ascii', 'ignore'))
self.write("</p>")
self.write("</body></html>")
def test_wordcount(get_text_mock, spark_context):
get_text_mock.return_value = "foo bar foo"
result = analyze(spark_context)
assert result[0] == ('foo', 2)
assert result[1] == ('bar', 1)
def analyze_file(self):
file = QtWidgets.QFileDialog.getOpenFileName(
self.MainWindow,
"Open a File",
filter="Wireshark capture file (*.pcap;*.pcapng);;All Files (*.*)")
self.df = analyze(file[0])
if self.df is not None:
self.ui.actionCreateGraph.setEnabled(True)
def handle_messages():
print "Handling Messages"
payload = request.get_data()
print payload
for sender, message in messaging_events(payload):
print "Incoming from %s: %s" % (sender, message)
send_message(PAT, sender, analyze(message))
return "ok"
def test(filename, graph):
total = []
data, features, labels = parse_csv(filename)
SGD_pred = runSGD(data)
total.append(analyze(SGD_pred, labels))
SVC_pred = runSVC(data)
total.append(analyze(SVC_pred, labels))
MNB_pred = runMNB(data)
total.append(analyze(MNB_pred, labels))
# each element in total looks like:
# [tn, fp, fn, tp, precision, recall, acc, fscore]
if (graph):
bargraph([i[0] for i in total], [i[1] for i in total], [i[2] for i in total], [i[3] for i in total], [i[4] for i in total], [i[5] for i in total], [i[6] for i in total], [i[7] for i in total])
def run(self):
print "Starting consumer thread"
a = analyze()
global queue
global tweetstore
global rtstore
block = ["insafediver", "MakeUseOf", "healthyworld24", "ISSAboveYou", "JoeGumby1", "Gumbletech"]
blockterms = ["invest", "market", "nasa", "untuk"]
prefuser = ["davewiner", "scobleizer"]
while True:
#lock.acquire()
condition.acquire()
if not queue:
#print "Nothing in queue: printer will try again"
condition.wait()
pass
else:
num = queue.pop()
screen_name = ""
text = ""
if num.has_key('user'):
if num['user'].has_key('screen_name'):
screen_name = num['user']['screen_name'].encode('ascii', 'ignore')
if num.has_key('text'):
text = num['text'].encode('ascii', 'ignore')
if screen_name in prefuser:
# print screen_name, text
for client in wsckts:
wsckts[client]['object'].send_msg(num)
greenstore.append(num)
# if self.client:
# self.db.tweets.insert_one(num)
elif text.startswith('RT'):
#rtstore.append(num)
pass
elif (text.find('http') != -1):
if (screen_name not in block):
blck = True
for term in blockterms:
if text.lower().find(term) > 0:
blck = False
if (blck) and not (a.cleanstr(text).isupper()):
if isgreen(num):
for client in wsckts:
wsckts[client]['object'].send_msg(num)
greenstore.append(num)
# if self.client:
# self.db.tweets.insert_one(num)
else:
for client in rwsckts:
rwsckts[client]['object'].send_msg(num)
tweetstore.append(num)
# if self.client:
# self.db2.redtweets.insert_one(num)
condition.release()
def add_entry():
db = get_db()
text = request.form['text']
analysis = analyze(text)
bytes = cPickle.dumps(analysis[1], 1)
db.execute('insert into entries (text, time, tones) values (?, ?, ?)',
[text, analysis[0], str(bytes)])
db.commit()
flash('New entry was successfully posted')
return redirect(url_for('analyzeWeb'))
def save_and_analyze_file(self):
file_name = QtWidgets.QFileDialog.getSaveFileName(
self.MainWindow,
"Save into a File",
filter="Wireshark capture file (*.pcap;*.pcapng);;All Files (*.*)")
if file_name[0]:
self.sniffer.write_into_pcap(file_path_name=file_name[0])
self.df = analyze(file_name[0])
if self.df is not None:
self.ui.actionCreateGraph.setEnabled(True)
def isgreen(tweet):
if tweet.has_key('text'):
text = tweet['text'].encode('ascii', 'ignore')
text = convertLinks(text)
if tweet.has_key('screen_name'):
screen_name = "@" + tweet['screen_name'].encode('ascii', 'ignore') + ": "
elif tweet.has_key('user'):
if tweet['user'].has_key('screen_name'):
screen_name = "@" + tweet['user']['screen_name'].encode('ascii', 'ignore') + ": "
a = analyze()
if a.finduser(screen_name) or (a.findword(text) or (a.findterm(text))):
return True
else:
return False
def get(self):
a = analyze()
header = """
<html>
<head>
<title>Green Tweets</Title>
<meta http-equiv="refresh" content="30" />
</head>
<body style="background:lightgray">
"""
self.write(header)
for tweet in tweetstore:
screen_name = ""
text = ""
if a.cleanstr(tweet['text']) == '':
#print "Skipping: " + tweet['text']
pass
else:
if tweet.has_key('text'):
text = tweet['text'].encode('ascii', 'ignore')
text = convertLinks(text)
if tweet.has_key('screen_name'):
screen_name = "@" + tweet['screen_name'].encode(
'ascii', 'ignore') + ": "
elif tweet.has_key('user'):
if tweet['user'].has_key('screen_name'):
screen_name = "@" + tweet['user'][
'screen_name'].encode('ascii', 'ignore') + ": "
if a.finduser(screen_name) or (a.findword(text)) or (
a.findterm(text)):
img_url = ""
if tweet['user'].has_key('profile_image_url'):
img_url = tweet['user']['profile_image_url']
self.write(
"<p style='background:white; color: green'>")
if img_url != "":
self.write("<img src='" + img_url +
"' style='float:left'>")
self.write("<a href='http://twitter.com/" +
screen_name[1:-2] + "' target='_blank'>" +
screen_name + "</a>")
self.write("<br/>")
self.write(text)
if tweet.has_key('created_at'):
self.write(
"<br/>" +
tweet['created_at'].encode('ascii', 'ignore'))
self.write("</p>")
self.write("</body></html>")
def get(self):
a = analyze()
header = """
<html>
<head>
<title>Twitter Client</Title>
<meta http-equiv="refresh" content="60" />
</head>
<body style="background:lightgray">
"""
self.write(header)
self.write("<a href='/green'>Green Tweets</A>")
self.write("<a href='/red'>Red Tweets</A>")
self.write("<a href='/stats'>Stats</A>")
self.write("<A href='/rt'>RTs</A>")
self.write("<A href='gjson'>Green Tweets JSON</A>")
self.write("</body></html>")
def get(self):
a = analyze()
header = """
<html>
<head>
<title>Green Tweets</Title>
"""
self.write(header)
finhdr = """
</head>
<body style="background:lightgray">
"""
self.write(finhdr)
for tweet in greenstore:
self.write(tweet)
self.write("\n")
self.write("</body></html>")
def get(self):
a = analyze()
header = """
<html>
<head>
<title>Twitter Client</Title>
<meta http-equiv="refresh" content="60" />
</head>
<body style="background:lightgray">
"""
self.write(header)
for tweet in tweetstore:
screen_name = ""
text = ""
if a.cleanstr(tweet['text']) == '':
#print "Skipping: " + tweet['text']
pass
else:
if tweet.has_key('text'):
text = tweet['text'].encode('ascii', 'ignore')
text = convertLinks(text)
if tweet.has_key('screen_name'):
screen_name = "@" + tweet['screen_name'].encode(
'ascii', 'ignore') + ": "
elif tweet.has_key('user'):
if tweet['user'].has_key('screen_name'):
screen_name = "@" + tweet['user'][
'screen_name'].encode('ascii', 'ignore') + ": "
if a.finduser(screen_name):
self.write("<p style='color: green'>")
elif a.findword(text) or a.findterm(text):
self.write("<p style='color: green'>")
txt = a.appendword(text)
if txt != "":
text = txt
else:
self.write("<p style='color: red'>")
self.write(screen_name)
self.write(text)
self.write("</p>")
self.write("</body></html>")
catDir = cat[0]+'_nT'+cat[1]+'_nW'+cat[2]+'_nB'+cat[3]
datahists = {}
bkghists = {}
sighists = {}
if len(sys.argv)>1: outDir=sys.argv[1]
else:
outDir = os.getcwd()
outDir+='/'+pfix
if not os.path.exists(outDir): os.system('mkdir '+outDir)
outDir+='/'+cutString
if not os.path.exists(outDir): os.system('mkdir '+outDir)
outDir+='/'+catDir
if not os.path.exists(outDir): os.system('mkdir '+outDir)
category = {'isEM':cat[0],'nttag':cat[1],'nWtag':cat[2],'nbtag':cat[3]}
for data in dataList:
datahists.update(analyze(tTreeData,data,cutList,isotrig,False,doJetRwt,iPlot,plotList[iPlot],category,region))
if catInd==nCats: del tFileData[data]
for bkg in bkgList:
bkghists.update(analyze(tTreeBkg,bkg,cutList,isotrig,doAllSys,doJetRwt,iPlot,plotList[iPlot],category,region))
if catInd==nCats: del tFileBkg[bkg]
if doAllSys and catInd==nCats:
for syst in shapesFiles:
for ud in ['Up','Down']: del tFileBkg[bkg+syst+ud]
for sig in sigList:
for decay in decays:
sighists.update(analyze(tTreeSig,sig+decay,cutList,isotrig,doAllSys,doJetRwt,iPlot,plotList[iPlot],category,region))
if catInd==nCats: del tFileSig[sig+decay]
if doAllSys and catInd==nCats:
for syst in shapesFiles:
for ud in ['Up','Down']: del tFileSig[sig+decay+syst+ud]
if doQ2sys:
def test_wordcount_analyze(_, __):
result = analyze(Context())
assert len(result) == 327
assert result[:6] == [('ut', 17), ('eu', 16), ('vel', 14), ('nec', 14), ('quis', 12), ('vitae', 12)]
from analyze import * import summary as sm pcaps = all_pcaps() local, remote, dir = next(pcaps) results = analyze(local, remote, dir) local_sender = results[0] local_sender_steady = results[1] remote_sender = results[2] remote_sender_steady = results[3] throughput_quantiles, rtt_quantiles, host, protocol, start_time, _ = results[0] rtt_quantiles
from mock import patch
from pysparkling import Context
from jobs.{{cookiecutter.job}} import analyze
@patch('jobs.{{cookiecutter.job}}.{{ cookiecutter.project.replace(' ', '').replace('-', '') }}Context.initalize_counter')
@patch('jobs.{{cookiecutter.job}}.{{ cookiecutter.project.replace(' ', '').replace('-', '') }}Context.inc_counter')
def test_{{cookiecutter.job}}_analyze(_, __):
result = analyze(Context())
assert len(result) == 327
outDir += '/' + pfix
if not os.path.exists(outDir): os.system('mkdir ' + outDir)
#outDir+='/'+cutString
if not os.path.exists(outDir): os.system('mkdir ' + outDir)
outDir += '/' + catDir
if not os.path.exists(outDir): os.system('mkdir ' + outDir)
category = {
'isEM': cat[0],
'nttag': cat[1],
'nWtag': cat[2],
'nbtag': cat[3],
'njets': cat[4]
}
for data in dataList:
datahists.update(
analyze(tTreeData, data, cutList, False, iPlot, plotList[iPlot],
category))
if catInd == nCats: del tFileData[data]
pickle.dump(datahists, open(outDir + '/datahists_' + iPlot + '.p', 'wb'))
catInd += 1
catInd = 1
for cat in catList:
if not runBkgs: break
catDir = cat[0] + '_nT' + cat[1] + '_nW' + cat[2] + '_nB' + cat[
3] + '_nJ' + cat[4]
catDir = catDir.replace('_nT0p',
'').replace('_nW0p',
'').replace('_nB0p',
'').replace('_nJ0p', '')
bkghists = {}
if len(sys.argv) > 1: outDir = sys.argv[1]
if not os.path.exists(outDir): os.system('mkdir ' + outDir)
outDir += '/' + catDir
if not os.path.exists(outDir): os.system('mkdir ' + outDir)
category = {
'isEM': cat[0],
'nttag': cat[1],
'nWtag': cat[2],
'nbtag': cat[3],
'njets': cat[4]
}
for data in dataList:
print "*****" * 20
print "*****" * 20
print "[data] : ", category, region, isCategorized
datahists.update(
analyze(tTreeData, data, data, cutList, False, doJetRwt, iPlot,
plotList[iPlot], category, region, isCategorized))
if catInd == nCats: del tFileData[data]
pickle.dump(datahists, open(outDir + '/datahists_' + iPlot + '.p', 'wb'))
catInd += 1
catInd = 1
for cat in catList:
if not runBkgs: break
if skip(cat[4], cat[3]) and isCategorized:
continue #DO YOU WANT TO HAVE THIS??
catDir = cat[0] + '_nT' + cat[1] + '_nW' + cat[2] + '_nB' + cat[
3] + '_nJ' + cat[4]
bkghists = {}
if len(sys.argv) > 1: outDir = sys.argv[1]
else:
outDir = os.getcwd()
catDir = cat[0]+'_nT'+cat[1]+'_nW'+cat[2]+'_nB'+cat[3]
datahists = {}
bkghists = {}
sighists = {}
if len(sys.argv)>1: outDir=sys.argv[1]
else:
outDir = os.getcwd()+'/'
outDir+=pfix
if not os.path.exists(outDir): os.system('mkdir '+outDir)
if not os.path.exists(outDir+'/'+cutString): os.system('mkdir '+outDir+'/'+cutString)
outDir+='/'+cutString
if not os.path.exists(outDir+'/'+catDir): os.system('mkdir '+outDir+'/'+catDir)
outDir+='/'+catDir
category = {'isEM':cat[0],'nttag':cat[1],'nWtag':cat[2],'nbtag':cat[3]}
for data in dataList:
datahists.update(analyze(tTreeData,data,cutList,False,iPlot,plotList[iPlot],category))
if catInd==nCats: del tFileData[data]
for bkg in bkgList:
bkghists.update(analyze(tTreeBkg,bkg,cutList,doAllSys,iPlot,plotList[iPlot],category))
if catInd==nCats: del tFileBkg[bkg]
if doAllSys and catInd==nCats:
for syst in shapesFiles:
if 'DataDriven' in bkg: continue
for ud in ['Up','Down']: del tFileBkg[bkg+syst+ud]
for sig in sigList:
for decay in decays:
sighists.update(analyze(tTreeSig,sig+decay,cutList,doAllSys,iPlot,plotList[iPlot],category))
if catInd==nCats: del tFileSig[sig+decay]
if doAllSys and catInd==nCats:
for syst in shapesFiles:
if 'DataDriven' in bkg: continue
import json
from analyze import *
from load_data import load_data
if __name__ == '__main__':
with open('keys.json') as f:
key = json.loads(f.read())['usda-api-key']
#load_data(key)
analyze()
outDir = os.getcwd()
outDir += '/' + pfix
if not os.path.exists(outDir): os.system('mkdir ' + outDir)
outDir += '/' + cutString
if not os.path.exists(outDir): os.system('mkdir ' + outDir)
outDir += '/' + catDir
if not os.path.exists(outDir): os.system('mkdir ' + outDir)
category = {
'isEM': cat[0],
'nttag': cat[1],
'nWtag': cat[2],
'nbtag': cat[3]
}
for data in dataList:
datahists.update(
analyze(tTreeData, data, cutList, isotrig, False, doJetRwt, iPlot,
plotList[iPlot], category, region))
if catInd == nCats: del tFileData[data]
for bkg in bkgList:
bkghists.update(
analyze(tTreeBkg, bkg, cutList, isotrig, doAllSys, doJetRwt, iPlot,
plotList[iPlot], category, region))
if catInd == nCats: del tFileBkg[bkg]
if doAllSys and catInd == nCats:
for syst in shapesFiles:
for ud in ['Up', 'Down']:
del tFileBkg[bkg + syst + ud]
for sig in sigList:
for decay in decays:
sighists.update(
analyze(tTreeSig, sig + decay, cutList, isotrig, doAllSys,
doJetRwt, iPlot, plotList[iPlot], category, region))
if len(syslist) > 1:
print('Systems:')
for sysname in sorted(syslist):
print(" " + sysname)
print()
# analyze the systems:
if len(syslist) == 1 and syslist[0] == '':
print('no system?')
usage()
sys.exit(1)
switches = (show_events, show_disk, show_details, show_filesystem, show_mem, show_ping)
for sysname in sorted(syslist):
analyze(sysname, allSystems, switches)
print()
##------------------------------------------------------------------------------
## pretty-print the resulting dictionary:
#print('allSystems:')
#print()
#pp.pprint(allSystems)
#print()
# ----------------------------------------------------------------------------
#
# ----------------------------------------------------------------------------
# EOF:
outfile = outfile)
return datafiles
if __name__ == '__main__':
parser = argparse.ArgumentParser(description="Run the benchmark tests")
parser.add_argument('--engines', type=str, nargs='+', default=['ode','pqp','fcl'],
help="The collision checkers to test")
args = parser.parse_args()
# Self collision
self_collision_data = run_self_collision(args.engines)
out_basename = os.path.join(package_path, 'results', 'self_collision')
analyze(self_collision_data, title='Self Collision', out_basename=out_basename)
# Empty envirnment
empty_collision_data = run_environment_collision(args.engines)
out_basename = os.path.join(package_path, 'results', 'empty_env_collision')
analyze(empty_collision_data, title='Empty Environment Collisions', out_basename=out_basename)
# PR kitchen
from catkin.find_in_workspaces import find_in_workspaces
kitchen_env = find_in_workspaces(
search_dirs=['share'],
project='pr_ordata',
path='data/kitchen/pr_kitchen.env.xml',
first_match_only=True)[0]
kitchen_collision_data = run_environment_collision(args.engines, kitchen_env, 'kitchen')
import timefrom gensim.models.word2vec import Word2Vecfrom Utils.string_utils import clean_strfrom Utils.file_utils import find_filesfrom analysis_pipeline import analyze, debug_analyzefrom analysis_pipeline import build_synonym_filter, fact_case_sensitive_stop_word_filter, fact_stop_word_filterfrom analysis_pipeline import fact_is_synonym_filter, white_space_tokenize, remove_punct_at_end_filter, lower_case_filter, remove_empty_tokens_filterfrom Config.train_word2vec_model_config import TrainWord2VecModelConfigimport sys
""" TRAIN Word 2 Vec Model"""
if len(sys.argv) != 2: raise Exception("Incorrect number of arguments passed - one expected, the config file name")
config = TrainWord2VecModelConfig(sys.argv[1])
""" Load analysis chain """syn_mapper = build_synonym_filter(config.keywords_files, config.case_sensitive)
if config.case_sensitive: stop_filter = fact_case_sensitive_stop_word_filter(config.stop_words_file)else: stop_filter = fact_stop_word_filter(config.stop_words_file)
# Simon Hughes: This is quite inefficient, as each function is applied in turn# resulting in multiple passes over the token stream. While not currently a# big performance bottleneck, could be much faster.# - TODO: use functional composition to speed upis_a_synonym_filter = fact_is_synonym_filter(syn_mapper)analysis_chain = [clean_str, white_space_tokenize, remove_punct_at_end_filter, lower_case_filter, stop_filter, syn_mapper.map_synonyms, remove_empty_tokens_filter] # is_a_synonym_filter] - Un-comment to just train on keywords.
#Test#rslt = debug_analyze("$150k as400 Sr.\ Java/j2ee and the C#.! developer. FIT \"HOT\" dev. -IBM's business, sql server management", analysis_chain)
""" Load Documents """start = time.time()
sentences = []files = find_files(config.processed_documents_folder, config.file_mask, True)print("%s files found in %s" % (len(files), config.processed_documents_folder))
documents = []for i, fname in enumerate(files): with open(fname) as f: contents = f.read() sentences.extend(contents.split("\n"))end = time.time()print("Loading %i sentences took %s seconds" % (len(sentences), str(end - start)))
""" Analyze - clean, tokenize, extract phrases """print("%i sentences to process" % len(sentences))
tokenized = []print("Tokenizing sentences")for i, sent in enumerate(sentences): tokens = analyze(sent, analysis_chain) if len(tokens) >= config.min_sentence_length_words: tokenized.append(tokens) if i % 100000 == 0: print(i)
""" Train Model """
start = time.time()
print("Training Model. This could take a while (10-60 mins for moderate collections). Get a coffee")model = Word2Vec(tokenized, iter=config.training_iterations, size=config.vector_size, window=config.window_size, min_count=config.min_word_count, workers=config.workers, sample=1e-5, hs=0, negative=20)model.save(config.model_file)end = time.time()print "Took %s seconds" % (end - start)
print()
# display list of systems we found:
syslist = list(allSystems)
if len(syslist) > 1:
print('Systems:')
for sysname in sorted(syslist):
print(" " + sysname)
print()
# analyze the systems:
if len(syslist) > 0:
for sysname in sorted(syslist):
analyze(sysname, allSystems)
print()
# close the output file if we had one:
if lclvars.outfile != None:
lclvars.outfile.close()
##------------------------------------------------------------------------------
## pretty-print the resulting dictionary:
#print('allSystems:')
#print()
#pp.pprint(allSystems)
#print()
# ----------------------------------------------------------------------------
