def export(self, filename):
""" Export the track data in the appropriate format for subsequent
analysis. Note that the HDF5 format is intended to store references to
objects, rather than the tracks themselves, so we need to deal with
that differently here...
TODO(arl): Make sure that we are working with an exisiting HDF5 file!
"""
# log the output
logger.info('Exporting {0:d} tracks to file...'.format(self.n_tracks))
if not filename.endswith('hdf5'):
utils.export(filename, self.tracks)
else:
utils.export_HDF(filename, self.refs, dummies=self.dummies)
def export():
piece = raw_input("Enter piece ID > ")
filename = raw_input("Enter a filename > ")
utils.export(generation.get()[int(piece)], filename)
print "File written"
# -*- coding: utf-8 -*- import utils usernames = ['BarackObama', 'billclinton', 'br', 'Cristiano', 'Ibra_official', 'JohnCleese', 'jeffzrebiecsun', 'kerrywashington', 'NatGeo', 'tnyCloseRead'] count = 200 for username in usernames: tweets = utils.cleanTweets(utils.getTweets(username, count - 1)) utils.export(username+"-tweets.txt", tweets, "w+") print(username, "has done")
# -*- coding: utf-8 -*-
import utils
usernames = ['sertaberener', 'DemetAkalin', 'hulyavsar', 'sertaberener', 'gulbenergen', 'MuratBoz', 'Niltakipte']
count = 3000
for username in usernames:
tweets = utils.cleanTweets(utils.getTweets(username, count))
utils.export("data/"+username+"-tweets.txt", tweets, "w")
def export(self, file):
vs = self.vs.cpu().clone()
vs -= self.translations[None, :]
vs *= self.scale
export(file, vs, self.faces)
r'^/src/device/io/*$', r'^/src/engine/interpreter/*$', r'^/src/memory/*$', r'^/include/*$', r'^/include/cpu/*$', r'^/include/device/*$', r'^/include/memory/*$', r'^/include/monitor/*$', r'^/include/rtl/*$', ] BLACK_LIST = [ r'/build/', r'/export/', r'/gen-expr/', r'/kvm-diff/', r'/recorder/', r'/.git/', r'mips32', r'riscv32', r'x86', r'/resource/bbl', r'/engine/rv64', r'/device/audio.c', r'/device/vga.c', r'/device/keyboard.c', r'runall.sh', ] export(WHITE_LIST, BLACK_LIST)
def test_export(self):
with utils.export({'number': 'JACOB'}) as path:
workbook = openpyxl.load_workbook(path)
sheet = workbook['JACOB']
assert sheet['C3'].internal_value == 'JACOB'
workbook.close()
def simulate():
args = utils.process_args(vars(utils.parser.parse_args()))
print(args)
Ns, densities, solvers, budgets, nsim, costType, verbose, loadPrev, standardize = args
result_dict = []
result_colnums_names = [
'N', 'Density', 'Solver', 'Budget', 'Cost', 'Time_avg', 'Time_sd',
'Sol_avg', 'Sol_sd'
]
total_simulations = utils.getTotalSimulation(
[Ns, densities, budgets, costType])
total_simulations *= nsim
progress = 0
loadPrev_outer = loadPrev
try:
for N in Ns:
for density in densities:
for budget in budgets:
for cost in costType:
sols = np.zeros((nsim, len(solvers)))
times = np.zeros((nsim, len(solvers)))
if loadPrev:
try:
print(
"\nLoading previously saved test instances..."
)
try:
update_cost = False
sims, new_budget = utils.load_saved_instance(
N, density, budget, cost)
except:
print("Need to update costs...")
update_cost = True
sims, new_budget = utils.load_saved_instance(
N, density, budget, None)
except:
print(
"Failed to load... Creating new instances..."
)
sims = []
loadPrev = False
else:
print("Creating new instances...")
sims = []
for sim in range(nsim):
if loadPrev and sim < len(sims):
changed_instance = False
G, B, U, C = sims[sim]
if update_cost:
print("\nUpdating costs...")
C = generate_cost(G, cost)
sims[sim] = G, B, U, C
changed_instance = True
if new_budget:
print(
"\nReusing test cases but with different budget..."
)
B = 5 * G.order() * budget
else:
changed_instance = True
G = generate_random_dag(N, density)
B = 5 * N * budget
U = generate_utility(G)
C = generate_cost(G, cost)
sims.append((G, B, U, C))
for solver_index in range(len(solvers)):
solver = solvers[solver_index]
if solver == "ilp":
if cost == "monotone":
C_ilp = C[0]
s_time, s_sol = ilp_time(G, C[0], B, U)
elif cost == "add":
s_time, s_sol = ilp_time(G, C, B, U)
elif solver == "bf":
s_time, s_sol = brute_force_time(
G, C, B, U, cost)
elif solver == "gd":
s_time, s_sol = greedy_time(
G, C, B, U, cost)
elif solver == "gd2":
s_time, s_sol = greedy2_time(
G, C, B, U, cost)
sols[sim, solver_index] = s_sol
times[sim, solver_index] = s_time
progress += 1
if verbose:
utils.update_progress(progress /
total_simulations)
if changed_instance or new_budget:
print("\nTest instances saved for future use.")
utils.save_instance(sims, N, density, budget, cost)
result_dict.extend(
utils.generate_result_dict(N, density, budget,
cost, solvers, sols,
times, standardize))
loadPrev = loadPrev_outer
utils.export(result_colnums_names, result_dict)
except KeyboardInterrupt:
utils.export(result_colnums_names, result_dict)
def conll(data,
cols=('form', 'postag', 'chunktag', 'guesstag'),
*args,
**kwargs):
"""Evaluates chunking f1-score provided with data with the following fields:
form, postag, chunktag, guesstag
Currently uses the CoNLL-2000 evaluation script to make the estimate.
This method will be deprecated with version 0.2
:param data: np.array
:param cols: columns to be used for the evaluation
:type cols: str or tuple or list
:return: f1-score estimate
:rtype: AccuracyResults
"""
warnings.warn(
'Using the CoNLL-2000 evaluation script is deprecated. `bio` '
'evaluation should be used instead.')
try:
os.makedirs(join(os.getcwd(), 'tmp/'))
except OSError:
pass
td = join(os.getcwd(), 'tmp/')
rn = rnd.randint(1000, 1000000000000)
fp_dp = join(td,
'chdata.%s.%s.tmp' % (time.asctime().replace(' ', ''), rn))
fp_res = join(td,
'chres.%s.%s.tmp' % (time.asctime().replace(' ', ''), rn))
fh_out = open(fp_res, 'w')
export(data, open(fp_dp, 'w'), cols=cols, ts=' ')
cwd = os.getcwd()
prl = join(cwd, 'conll_eval.pl' + random_str())
with open(prl, 'w') as fh:
fh.write(conll_script)
c = cmd('perl %s -l < {}' % prl, fp_dp, cwd=cwd, stdout=fh_out)
r = AccuracyResults()
try:
check_call(c)
r.parse_conll_eval_table(fp_res)
except CalledProcessError:
exc_type, exc_value, exc_traceback = sys.exc_info()
print "*** print_tb:"
traceback.print_tb(exc_traceback, limit=1, file=sys.stdout)
print "*** print_exception:"
traceback.print_exception(exc_type,
exc_value,
exc_traceback,
limit=2,
file=sys.stdout)
finally:
os.remove(fp_dp)
os.remove(fp_res)
os.remove(prl)
return r
def conll(data, cols=('form', 'postag', 'chunktag', 'guesstag'),
*args, **kwargs):
"""Evaluates chunking f1-score provided with data with the following fields:
form, postag, chunktag, guesstag
Currently uses the CoNLL-2000 evaluation script to make the estimate.
This method will be deprecated with version 0.2
:param data: np.array
:param cols: columns to be used for the evaluation
:type cols: str or tuple or list
:return: f1-score estimate
:rtype: AccuracyResults
"""
warnings.warn('Using the CoNLL-2000 evaluation script is deprecated. `bio` '
'evaluation should be used instead.')
try:
os.makedirs(join(os.getcwd(), 'tmp/'))
except OSError:
pass
td = join(os.getcwd(), 'tmp/')
rn = rnd.randint(1000, 1000000000000)
fp_dp = join(td, 'chdata.%s.%s.tmp' % (time.asctime().replace(' ', ''), rn))
fp_res = join(td, 'chres.%s.%s.tmp' % (time.asctime().replace(' ', ''), rn))
fh_out = open(fp_res, 'w')
export(data,
open(fp_dp, 'w'),
cols=cols,
ts=' ')
cwd = os.getcwd()
prl = join(cwd, 'conll_eval.pl' + random_str())
with open(prl, 'w') as fh:
fh.write(conll_script)
c = cmd(
'perl %s -l < {}' % prl,
fp_dp,
cwd=cwd,
stdout=fh_out
)
r = AccuracyResults()
try:
check_call(c)
r.parse_conll_eval_table(fp_res)
except CalledProcessError:
exc_type, exc_value, exc_traceback = sys.exc_info()
print "*** print_tb:"
traceback.print_tb(exc_traceback,
limit=1,
file=sys.stdout)
print "*** print_exception:"
traceback.print_exception(exc_type,
exc_value,
exc_traceback,
limit=2,
file=sys.stdout)
finally:
os.remove(fp_dp)
os.remove(fp_res)
os.remove(prl)
return r
def export(_id):
collection = mongo.db.courses
_id = bson.ObjectId(_id)
course = collection.find_one(_id)
with utils.export(course) as path:
return flask.send_file(path, mimetype='application/vnd.ms-excel', as_attachment=True)
# -*- coding: utf-8 -*-
import utils
import os
import nltk
positives, negatives = [], []
for filename in os.listdir("data"):
if filename.endswith(".txt"):
with open('data/' + filename) as f:
tweets = [tweet for tweet in f.readlines()]
pos, neg = utils.groupTweets(tweets)
utils.export("train/positives.txt", pos, "a")
utils.export("train/negatives.txt", neg, "a")
