def test_try_to_move_back_in_time(self):
# Create a nice event to make time fly
event = Event(10)
event.perform(self.world)
# Create an event that tries to move back in time
wrong_event = Event(5)
self.assertRaises(ValueError, wrong_event.perform, self.world)
def JSONtoEvent(j: json):
event = Event(j["min"], j["max"], j["Location"], j["Time"], j["Title"],
j["Description"])
persons = []
for participant in j["Participants"]:
persons.append(
Participant(participant[0]["Email"], participant[0]["Name"]))
event.addParticipants(persons)
return event
def main():
granularity = 15 # TODO allow change of granularity
dt_start = sys.argv[1]
dt_end = sys.argv[2]
tx_location = sys.argv[3]
tx_rrule = sys.argv[4]
temp = open("C:/wamp/www/mesa/python/temp1.json",
"r") # TODO make relative path
bl_calendars = json.loads(temp.read())
temp.close()
temp = open("C:/wamp/www/mesa/python/temp2.json", "r")
bl_settings = json.loads(temp.read())
temp.close()
rrule = parse_rrule(tx_rrule) # TODO create use for rrule
priorities = parse_priorities(bl_settings)
original_event = Event(
"blevent", {
"blEvent": {
"start_time": dt_start.replace(" ", "T") + "Z",
"end_time": dt_end.replace(" ", "T") + "Z",
"location": tx_location,
"travel_time": 0
}
})
calendar_set = construct_calendar_set(bl_calendars)
point_list = construct_point_list(calendar_set, granularity,
original_event, bl_settings)
cost_output = smallest_cost(point_list, priorities, original_event,
granularity, tx_location, calendar_set)
print(cost_output)
def test_add_event(self):
shuffled_events = [Event(time) for time in range(10, 41, 10)]
random.shuffle(shuffled_events)
for event in shuffled_events:
self.world.add_event(event)
self.assertSequenceEqual([event.time for event in self.world.stack],
range(10, 41, 10))
def find_events():
events_list = []
try:
events_url = "https://api.betsapi.com/v2/events/upcoming?token={}&sport_id=151?day=TODAY".format(
os.environ.get("BETS_API_TOKEN"))
events = requests.get(events_url).json()
current_epoch_time = int(time.time())
events_access = events["results"]
except:
logging.info("Couldnt get upcoming events")
else:
for events_data in events_access:
try:
event_id = events_data["id"]
event_time = events_data["time"]
home_team = events_data["home"]["name"]
away_team = events_data["away"]["name"]
league = events_data["league"]["name"]
except:
logging.debug(
"Could not retrieve event id and time in find_events()")
else:
# get events within the next 2 hours (7200 sec)
if (int(event_time) - int(current_epoch_time)) > 0 and (
int(event_time) - int(current_epoch_time)) < 7200:
events_list.append(
Event(event_id, event_time, home_team, away_team,
league))
# if we find a good event/bet add it
for event in events_list:
try:
oddsUrl = "https://api.betsapi.com/v2/event/odds?token={}&event_id={}&odds_market=1&source=pinnaclesports".format(
os.environ.get("BETS_API_TOKEN"), event.eventId)
odds = requests.get(oddsUrl).json()
odds_access = odds["results"]["odds"]["151_1"]
homeOddsNow = float(odds_access[0]["home_od"])
homeOddsEarlier = float(odds_access[1]["home_od"])
awayOddsNow = float(odds_access[0]["away_od"])
awayOddsEarlier = float(odds_access[1]["away_od"])
if homeOddsNow < (homeOddsEarlier * 0.95):
if do_bet_exist(event.eventId) == False:
place_bet(event.eventId, "H", homeOddsNow, 50)
logging.info(
"Placed bet on home team {} vs {} with id {}".format(
event.home_team, event.away_team, event.eventId))
if awayOddsNow < (awayOddsEarlier * 0.95):
if do_bet_exist(event.eventId) == False:
place_bet(event.eventId, "A", awayOddsNow, 50)
logging.info(
"Placed bet on away team {} vs {} with id {}".format(
event.home_team, event.away_team, event.eventId))
except:
logging.debug("No event odds in find_events()")
def endElement(self, tag):
if tag == "trace":
if self.log in self.logs:
if self.tracename in self.logs[self.log]:
print(self.log)
print(self.tracename)
raise RuntimeError("Trace already exists")
else:
self.logs[self.log] = {}
self.start_events[self.log] = collections.defaultdict(int)
self.end_events[self.log] = collections.defaultdict(int)
if (self.lifecycle_options['lifecycle_exists'] == 0):
self.start_events[self.log][self.event_start] += 1
self.end_events[self.log][self.event_end] += 1
elif (self.lifecycle_options['lifecycle_exists'] == 1
and len(self.lifecycle_options['lifecycle_options']) == 1):
self.start_events[self.log][self.event_start] += 1
self.end_events[self.log][self.event_end] += 1
else:
self.start_events[self.log][self.event_start_start] += 1
self.end_events[self.log][self.event_end_start] += 1
self.logs[self.log][self.tracename] = Trace(
self.log, self.tracename, self.trace)
self.event_start = None
self.event_end = None
self.event_start_start = None
self.event_end_start = None
self.tracename = None
self.trace = []
if tag == "event":
self.event_map[self.attrib['concept:name']] += 1
if 'lifecycle:transition' not in self.attrib:
self.lifecycle_options[
'lifecycle_exists'] = 0 # if there is at least one event without a lifecycle transition, then we cannot use lifecycles.
else:
self.lifecycle_options['lifecycle_options'].add(
self.attrib['lifecycle:transition'])
if self.lifecycle_options[
'lifecycle_exists'] == 0 or self.lifecycle_options[
'lifecycle_exists'] == 1 and self.attrib[
'lifecycle:transition'] == "start":
if not self.event_start_start:
self.event_start_start = self.attrib['concept:name']
self.event_end_start = self.attrib['concept:name']
if self.lifecycle_options[
'lifecycle_exists'] == 0 or self.lifecycle_options[
'lifecycle_exists'] == 1 and self.attrib[
'lifecycle:transition'] == "complete":
if not self.event_start:
self.event_start = self.attrib['concept:name']
self.event_end = self.attrib['concept:name']
self.trace.append(Event(self.log, self.tracename, self.attrib))
self.attrib = None
self.tag = tag
def get_all_events(self,
calendar_id: int,
strip_private=False) -> List[Event]:
''' Return list of Event objects that match the provided calendar_id.
If strip_private is True, all non-private events will be returned. '''
cid: Tuple = (calendar_id, )
if strip_private:
self._execute(
"SELECT * FROM events WHERE calendar_id = ? AND private = 0",
cid)
else:
self._execute("SELECT * FROM events WHERE calendar_id = ?", cid)
results = self._get_all_results()
return [Event(*event)
for event in results] # list comps are so comfy unf
def build_workout_sequence(self, duration, profile):
# determine workout format
exercise_sequence = []
if self.profile['warmup']:
duration -= 2
exercise_sequence.append(Event(self.make_exercise('Warmup'), 120))
total_rounds = int(np.floor(duration * 60 / 200))
total_exercises = total_rounds * 2
exercises_list = self.do_sampling(profile, total_exercises)
def divide_chunks(l, n):
for i in range(0, len(l), n):
yield l[i:i + n]
num_humans = 1 if 'num_humans' not in profile else profile['num_humans']
#this is reaaly only a 2 human format
chunksize = 2
for chunk in divide_chunks(exercises_list, chunksize):
if len(chunk) < chunksize:
continue #non multiple number of minutes
if num_humans == 1:
for iter in range(2):
exercise_sequence.append(Event(chunk[0], 40))
exercise_sequence.append(Event(chunk[1], 40))
exercise_sequence.append(
Event(self.make_exercise('Rest'), 20))
else:
for iter in range(2):
exercise_sequence.append(Event((chunk[0], chunk[1]), 40))
exercise_sequence.append(Event((chunk[1], chunk[0]), 40))
exercise_sequence.append(
Event(self.make_exercise('Rest'), 20))
exercise_sequence = exercise_sequence[:-1] #remove final rest
# #30 reps of each exercise; do full circuit twice
return exercise_sequence
def get_event(self, event_id: int) -> Event:
''' Return a single event, selected by event_id. '''
eid: Tuple = (event_id, )
self._execute("SELECT * FROM events WHERE event_id = ?", eid)
return Event(*self._get_result())
def project(XK, XV, LorY, surfSrc, surfTar, K_diag, V_diag, IorE, self, param,
ind0, timing, kernel):
if param.GPU == 1:
tic = cuda.Event()
toc = cuda.Event()
else:
tic = Event()
toc = Event()
REAL = param.REAL
Ns = len(surfSrc.triangle)
Nt = len(surfTar.triangle)
L = numpy.sqrt(2 * surfSrc.Area) # Representative length
tic.record()
K = param.K
w = getWeights(K)
X_V = numpy.zeros(Ns * K)
X_Kx = numpy.zeros(Ns * K)
X_Ky = numpy.zeros(Ns * K)
X_Kz = numpy.zeros(Ns * K)
X_Kc = numpy.zeros(Ns * K)
X_Vc = numpy.zeros(Ns * K)
NsK = numpy.arange(Ns * K)
X_V[:] = XV[NsK / K] * w[NsK % K] * surfSrc.Area[NsK / K]
X_Kx[:] = XK[NsK / K] * w[NsK % K] * surfSrc.Area[NsK /
K] * surfSrc.normal[NsK /
K, 0]
X_Ky[:] = XK[NsK / K] * w[NsK % K] * surfSrc.Area[NsK /
K] * surfSrc.normal[NsK /
K, 1]
X_Kz[:] = XK[NsK / K] * w[NsK % K] * surfSrc.Area[NsK /
K] * surfSrc.normal[NsK /
K, 2]
X_Kc[:] = XK[NsK / K]
X_Vc[:] = XV[NsK / K]
toc.record()
toc.synchronize()
timing.time_mass += tic.time_till(toc) * 1e-3
tic.record()
C = 0
getMultipole(surfSrc.tree, C, surfSrc.xj, surfSrc.yj, surfSrc.zj, X_V,
X_Kx, X_Ky, X_Kz, ind0, param.P, param.NCRIT)
toc.record()
toc.synchronize()
timing.time_P2M += tic.time_till(toc) * 1e-3
tic.record()
for C in reversed(range(1, len(surfSrc.tree))):
PC = surfSrc.tree[C].parent
upwardSweep(surfSrc.tree, C, PC, param.P, ind0.II, ind0.JJ, ind0.KK,
ind0.index, ind0.combII, ind0.combJJ, ind0.combKK,
ind0.IImii, ind0.JJmjj, ind0.KKmkk, ind0.index_small,
ind0.index_ptr)
toc.record()
toc.synchronize()
timing.time_M2M += tic.time_till(toc) * 1e-3
tic.record()
X_V = X_V[surfSrc.sortSource]
X_Kx = X_Kx[surfSrc.sortSource]
X_Ky = X_Ky[surfSrc.sortSource]
X_Kz = X_Kz[surfSrc.sortSource]
X_Kc = X_Kc[surfSrc.sortSource]
X_Vc = X_Vc[surfSrc.sortSource]
toc.record()
toc.synchronize()
timing.time_sort += tic.time_till(toc) * 1e-3
param.Nround = len(surfTar.twig) * param.NCRIT
K_aux = numpy.zeros(param.Nround)
V_aux = numpy.zeros(param.Nround)
AI_int = 0
### CPU code
if param.GPU == 0:
K_aux, V_aux = M2P_sort(surfSrc, surfTar, K_aux, V_aux, self,
ind0.index_large, param, LorY, timing)
K_aux, V_aux = P2P_sort(surfSrc, surfTar, X_V, X_Kx, X_Ky, X_Kz, X_Kc,
X_Vc, K_aux, V_aux, self, LorY, K_diag, V_diag,
IorE, L, w, param, timing)
### GPU code
elif param.GPU == 1:
K_gpu = cuda.to_device(K_aux.astype(REAL))
V_gpu = cuda.to_device(V_aux.astype(REAL))
if surfTar.offsetMlt[self, len(surfTar.twig)] > 0:
K_gpu, V_gpu = M2P_gpu(surfSrc, surfTar, K_gpu, V_gpu, self, ind0,
param, LorY, timing, kernel)
K_gpu, V_gpu = P2P_gpu(surfSrc, surfTar, X_V, X_Kx, X_Ky, X_Kz, X_Kc,
X_Vc, K_gpu, V_gpu, self, LorY, K_diag, IorE, L,
w, param, timing, kernel)
tic.record()
K_aux = cuda.from_device(K_gpu, len(K_aux), dtype=REAL)
V_aux = cuda.from_device(V_gpu, len(V_aux), dtype=REAL)
toc.record()
toc.synchronize()
timing.time_trans += tic.time_till(toc) * 1e-3
tic.record()
K_lyr = K_aux[surfTar.unsort]
V_lyr = V_aux[surfTar.unsort]
toc.record()
toc.synchronize()
timing.time_sort += tic.time_till(toc) * 1e-3
return K_lyr, V_lyr
def project_Kt(XKt, LorY, surfSrc, surfTar, Kt_diag, self, param, ind0, timing,
kernel):
if param.GPU == 1:
tic = cuda.Event()
toc = cuda.Event()
else:
tic = Event()
toc = Event()
REAL = param.REAL
Ns = len(surfSrc.triangle)
Nt = len(surfTar.triangle)
L = numpy.sqrt(2 * surfSrc.Area) # Representative length
tic.record()
K = param.K
w = getWeights(K)
X_Kt = numpy.zeros(Ns * K)
X_Ktc = numpy.zeros(Ns * K)
NsK = numpy.arange(Ns * K)
X_Kt[:] = XKt[NsK / K] * w[NsK % K] * surfSrc.Area[NsK / K]
X_Ktc[:] = XKt[NsK / K]
toc.record()
toc.synchronize()
timing.time_mass += tic.time_till(toc) * 1e-3
tic.record()
C = 0
X_aux = numpy.zeros(Ns * K)
getMultipole(surfSrc.tree, C, surfSrc.xj, surfSrc.yj, surfSrc.zj, X_Kt,
X_aux, X_aux, X_aux, ind0, param.P, param.NCRIT)
toc.record()
toc.synchronize()
timing.time_P2M += tic.time_till(toc) * 1e-3
tic.record()
for C in reversed(range(1, len(surfSrc.tree))):
PC = surfSrc.tree[C].parent
upwardSweep(surfSrc.tree, C, PC, param.P, ind0.II, ind0.JJ, ind0.KK,
ind0.index, ind0.combII, ind0.combJJ, ind0.combKK,
ind0.IImii, ind0.JJmjj, ind0.KKmkk, ind0.index_small,
ind0.index_ptr)
toc.record()
toc.synchronize()
timing.time_M2M += tic.time_till(toc) * 1e-3
tic.record()
X_Kt = X_Kt[surfSrc.sortSource]
X_Ktc = X_Ktc[surfSrc.sortSource]
toc.record()
toc.synchronize()
timing.time_sort += tic.time_till(toc) * 1e-3
param.Nround = len(surfTar.twig) * param.NCRIT
Ktx_aux = numpy.zeros(param.Nround)
Kty_aux = numpy.zeros(param.Nround)
Ktz_aux = numpy.zeros(param.Nround)
AI_int = 0
### CPU code
if param.GPU == 0:
if surfTar.offsetMlt[self, len(surfTar.twig)] > 0:
Ktx_aux, Kty_aux, Ktz_aux = M2PKt_sort(surfSrc, surfTar, Ktx_aux,
Kty_aux, Ktz_aux, self,
ind0.index_large, param,
LorY, timing)
Ktx_aux, Kty_aux, Ktz_aux = P2PKt_sort(surfSrc, surfTar, X_Kt, X_Ktc,
Ktx_aux, Kty_aux, Ktz_aux, self,
LorY, w, param, timing)
### GPU code
elif param.GPU == 1:
Ktx_gpu = cuda.to_device(Ktx_aux.astype(REAL))
Kty_gpu = cuda.to_device(Kty_aux.astype(REAL))
Ktz_gpu = cuda.to_device(Ktz_aux.astype(REAL))
if surfTar.offsetMlt[self, len(surfTar.twig)] > 0:
Ktx_gpu, Kty_gpu, Ktz_gpu = M2PKt_gpu(surfSrc, surfTar, Ktx_gpu,
Kty_gpu, Ktz_gpu, self, ind0,
param, LorY, timing, kernel)
Ktx_gpu, Kty_gpu, Ktz_gpu = P2PKt_gpu(surfSrc, surfTar, X_Kt, X_Ktc,
Ktx_gpu, Kty_gpu, Ktz_gpu, self,
LorY, w, param, timing, kernel)
tic.record()
Ktx_aux = cuda.from_device(Ktx_gpu, len(Ktx_aux), dtype=REAL)
Kty_aux = cuda.from_device(Kty_gpu, len(Kty_aux), dtype=REAL)
Ktz_aux = cuda.from_device(Ktz_gpu, len(Ktz_aux), dtype=REAL)
toc.record()
toc.synchronize()
timing.time_trans += tic.time_till(toc) * 1e-3
tic.record()
Kt_lyr = Ktx_aux[surfTar.unsort]*surfTar.normal[:,0] \
+ Kty_aux[surfTar.unsort]*surfTar.normal[:,1] \
+ Ktz_aux[surfTar.unsort]*surfTar.normal[:,2]
if abs(Kt_diag) > 1e-12: # if same surface
Kt_lyr += Kt_diag * XKt
toc.record()
toc.synchronize()
timing.time_sort += tic.time_till(toc) * 1e-3
return Kt_lyr
def test_run(self):
self.world.stack = [Event(time) for time in range(10, 41, 10)]
self.world.run()
def project(XK, XV, LorY, surfSrc, surfTar, K_diag, V_diag, IorE,
self, param, ind0, timing, kernel):
if param.GPU==1:
tic = cuda.Event()
toc = cuda.Event()
else:
tic = Event()
toc = Event()
REAL = param.REAL
Ns = len(surfSrc.triangle)
Nt = len(surfTar.triangle)
L = numpy.sqrt(2*surfSrc.Area) # Representative length
tic.record()
K = param.K
w = getWeights(K)
X_V = numpy.zeros(Ns*K)
X_Kx = numpy.zeros(Ns*K)
X_Ky = numpy.zeros(Ns*K)
X_Kz = numpy.zeros(Ns*K)
X_Kc = numpy.zeros(Ns*K)
X_Vc = numpy.zeros(Ns*K)
NsK = numpy.arange(Ns*K)
X_V[:] = XV[NsK/K]*w[NsK%K]*surfSrc.Area[NsK/K]
X_Kx[:] = XK[NsK/K]*w[NsK%K]*surfSrc.Area[NsK/K]*surfSrc.normal[NsK/K,0]
X_Ky[:] = XK[NsK/K]*w[NsK%K]*surfSrc.Area[NsK/K]*surfSrc.normal[NsK/K,1]
X_Kz[:] = XK[NsK/K]*w[NsK%K]*surfSrc.Area[NsK/K]*surfSrc.normal[NsK/K,2]
X_Kc[:] = XK[NsK/K]
X_Vc[:] = XV[NsK/K]
toc.record()
toc.synchronize()
timing.time_mass += tic.time_till(toc)*1e-3
tic.record()
C = 0
getMultipole(surfSrc.tree, C, surfSrc.xj, surfSrc.yj, surfSrc.zj,
X_V, X_Kx, X_Ky, X_Kz, ind0, param.P, param.NCRIT)
toc.record()
toc.synchronize()
timing.time_P2M += tic.time_till(toc)*1e-3
tic.record()
for C in reversed(range(1,len(surfSrc.tree))):
PC = surfSrc.tree[C].parent
upwardSweep(surfSrc.tree, C, PC, param.P, ind0.II, ind0.JJ, ind0.KK, ind0.index, ind0.combII, ind0.combJJ,
ind0.combKK, ind0.IImii, ind0.JJmjj, ind0.KKmkk, ind0.index_small, ind0.index_ptr)
toc.record()
toc.synchronize()
timing.time_M2M += tic.time_till(toc)*1e-3
tic.record()
X_V = X_V[surfSrc.sortSource]
X_Kx = X_Kx[surfSrc.sortSource]
X_Ky = X_Ky[surfSrc.sortSource]
X_Kz = X_Kz[surfSrc.sortSource]
X_Kc = X_Kc[surfSrc.sortSource]
X_Vc = X_Vc[surfSrc.sortSource]
toc.record()
toc.synchronize()
timing.time_sort += tic.time_till(toc)*1e-3
param.Nround = len(surfTar.twig)*param.NCRIT
K_aux = numpy.zeros(param.Nround)
V_aux = numpy.zeros(param.Nround)
AI_int = 0
### CPU code
if param.GPU==0:
K_aux, V_aux = M2P_sort(surfSrc, surfTar, K_aux, V_aux, self,
ind0.index_large, param, LorY, timing)
K_aux, V_aux = P2P_sort(surfSrc, surfTar, X_V, X_Kx, X_Ky, X_Kz, X_Kc, X_Vc,
K_aux, V_aux, self, LorY, K_diag, V_diag, IorE, L, w, param, timing)
### GPU code
elif param.GPU==1:
K_gpu = cuda.to_device(K_aux.astype(REAL))
V_gpu = cuda.to_device(V_aux.astype(REAL))
if surfTar.offsetMlt[self,len(surfTar.twig)]>0:
K_gpu, V_gpu = M2P_gpu(surfSrc, surfTar, K_gpu, V_gpu, self,
ind0, param, LorY, timing, kernel)
K_gpu, V_gpu = P2P_gpu(surfSrc, surfTar, X_V, X_Kx, X_Ky, X_Kz, X_Kc, X_Vc,
K_gpu, V_gpu, self, LorY, K_diag, IorE, L, w, param, timing, kernel)
tic.record()
K_aux = cuda.from_device(K_gpu, len(K_aux), dtype=REAL)
V_aux = cuda.from_device(V_gpu, len(V_aux), dtype=REAL)
toc.record()
toc.synchronize()
timing.time_trans += tic.time_till(toc)*1e-3
tic.record()
K_lyr = K_aux[surfTar.unsort]
V_lyr = V_aux[surfTar.unsort]
toc.record()
toc.synchronize()
timing.time_sort += tic.time_till(toc)*1e-3
return K_lyr, V_lyr
def project_Kt(XKt, LorY, surfSrc, surfTar, Kt_diag,
self, param, ind0, timing, kernel):
if param.GPU==1:
tic = cuda.Event()
toc = cuda.Event()
else:
tic = Event()
toc = Event()
REAL = param.REAL
Ns = len(surfSrc.triangle)
Nt = len(surfTar.triangle)
L = numpy.sqrt(2*surfSrc.Area) # Representative length
tic.record()
K = param.K
w = getWeights(K)
X_Kt = numpy.zeros(Ns*K)
X_Ktc = numpy.zeros(Ns*K)
NsK = numpy.arange(Ns*K)
X_Kt[:] = XKt[NsK/K]*w[NsK%K]*surfSrc.Area[NsK/K]
X_Ktc[:] = XKt[NsK/K]
toc.record()
toc.synchronize()
timing.time_mass += tic.time_till(toc)*1e-3
tic.record()
C = 0
X_aux = numpy.zeros(Ns*K)
getMultipole(surfSrc.tree, C, surfSrc.xj, surfSrc.yj, surfSrc.zj,
X_Kt, X_aux, X_aux, X_aux, ind0, param.P, param.NCRIT)
toc.record()
toc.synchronize()
timing.time_P2M += tic.time_till(toc)*1e-3
tic.record()
for C in reversed(range(1,len(surfSrc.tree))):
PC = surfSrc.tree[C].parent
upwardSweep(surfSrc.tree, C, PC, param.P, ind0.II, ind0.JJ, ind0.KK, ind0.index, ind0.combII, ind0.combJJ,
ind0.combKK, ind0.IImii, ind0.JJmjj, ind0.KKmkk, ind0.index_small, ind0.index_ptr)
toc.record()
toc.synchronize()
timing.time_M2M += tic.time_till(toc)*1e-3
tic.record()
X_Kt = X_Kt[surfSrc.sortSource]
X_Ktc = X_Ktc[surfSrc.sortSource]
toc.record()
toc.synchronize()
timing.time_sort += tic.time_till(toc)*1e-3
param.Nround = len(surfTar.twig)*param.NCRIT
Ktx_aux = numpy.zeros(param.Nround)
Kty_aux = numpy.zeros(param.Nround)
Ktz_aux = numpy.zeros(param.Nround)
AI_int = 0
### CPU code
if param.GPU==0:
if surfTar.offsetMlt[self,len(surfTar.twig)]>0:
Ktx_aux, Kty_aux, Ktz_aux = M2PKt_sort(surfSrc, surfTar, Ktx_aux, Kty_aux, Ktz_aux, self,
ind0.index_large, param, LorY, timing)
Ktx_aux, Kty_aux, Ktz_aux = P2PKt_sort(surfSrc, surfTar, X_Kt, X_Ktc,
Ktx_aux, Kty_aux, Ktz_aux, self, LorY, w, param, timing)
### GPU code
elif param.GPU==1:
Ktx_gpu = cuda.to_device(Ktx_aux.astype(REAL))
Kty_gpu = cuda.to_device(Kty_aux.astype(REAL))
Ktz_gpu = cuda.to_device(Ktz_aux.astype(REAL))
if surfTar.offsetMlt[self,len(surfTar.twig)]>0:
Ktx_gpu, Kty_gpu, Ktz_gpu = M2PKt_gpu(surfSrc, surfTar,
Ktx_gpu, Kty_gpu, Ktz_gpu, self,
ind0, param, LorY, timing, kernel)
Ktx_gpu, Kty_gpu, Ktz_gpu = P2PKt_gpu(surfSrc, surfTar, X_Kt, X_Ktc, Ktx_gpu, Kty_gpu, Ktz_gpu,
self, LorY, w, param, timing, kernel)
tic.record()
Ktx_aux = cuda.from_device(Ktx_gpu, len(Ktx_aux), dtype=REAL)
Kty_aux = cuda.from_device(Kty_gpu, len(Kty_aux), dtype=REAL)
Ktz_aux = cuda.from_device(Ktz_gpu, len(Ktz_aux), dtype=REAL)
toc.record()
toc.synchronize()
timing.time_trans += tic.time_till(toc)*1e-3
tic.record()
Kt_lyr = Ktx_aux[surfTar.unsort]*surfTar.normal[:,0] \
+ Kty_aux[surfTar.unsort]*surfTar.normal[:,1] \
+ Ktz_aux[surfTar.unsort]*surfTar.normal[:,2]
if abs(Kt_diag)>1e-12: # if same surface
Kt_lyr += Kt_diag * XKt
toc.record()
toc.synchronize()
timing.time_sort += tic.time_till(toc)*1e-3
return Kt_lyr
# Read sensor data from Arduino via serial
reading = ser.readline()
if ser.isButtonPressed(reading):
ongoingMeal.forceToggle(buffer.latest())
else:
if not buffer.isFull():
buffer.add(reading)
continue
outlier = buffer.score(reading,
useTypical=eventInProgress) > OUTLIER_Z_SCORE
if not eventInProgress:
buffer.updateTypicalDev()
if not eventInProgress and outlier:
eventInProgress = True
events.append(Event([buffer.earliest()], reading))
elif eventInProgress:
if outlier:
events[-1].add(reading)
else:
eventInProgress = False
events[-1].end()
ongoingMeal.updateWithEvent(events[-1])
buffer.add(reading)
if ongoingMeal.checkIfIdle() or ongoingMeal.forceEnd:
success = ongoingMeal.endMeal(buffer.latest())
if success == -1:
print("Invalid Meal!")
else:
def parse_file(folder):
print("######################")
print("PROCESSING: ")
print(folder)
print("######################")
list_events = []
folder = os.path.join(os.path.dirname(__file__), folder)
for filename in os.listdir(folder):
print(filename)
if "STAKING_FLEX" in filename:
filename = os.path.join(folder, filename)
with open(filename, "r") as a_file:
for line in a_file:
if not line.startswith("--"):
line_parts = re.split(r'\t+', line)
e = Event(trans_type="Staking",
buy_amt=line_parts[3],
buy_cur=line_parts[1],
comment="Binance Flexible Staking",
effective_date=parse_date(
line_parts[0], '%Y-%m-%d %H:%M:%S'),
exchange="Binance")
list_events.append(e)
if "STAKING_LOCKED" in filename:
filename = os.path.join(folder, filename)
with open(filename, "r") as a_file:
for line in a_file:
if not line.startswith("--"):
try:
line_parts = re.split(r'\t+', line)
e = Event(trans_type="Staking",
buy_amt=line_parts[3],
buy_cur=line_parts[0],
comment="Binance Locked Staking",
effective_date=parse_date(
line_parts[4], '%Y-%m-%d'),
exchange="Binance")
list_events.append(e)
except:
print("ERROR:")
print(line)
if "Buy History" in filename:
filename = os.path.join(folder, filename)
with open(filename, "r") as a_file:
df = pd.read_excel(filename, sheet_name="sheet1")
df = df[df["Status"] == "Completed"]
for index, row in df.iterrows():
e = Event(trans_type="Trade",
buy_amt=row["Final Amount"].split(" ")[0],
buy_cur=row["Final Amount"].split(" ")[1],
sell_amt=row["Amount"].split(" ")[0],
sell_cur=row["Amount"].split(" ")[1],
fee_amt=row["Fees"].split(" ")[0],
fee_cur=row["Fees"].split(" ")[1],
comment="Binance Buy",
effective_date=parse_date(
row["Date(UTC)"], '%Y-%m-%d %H:%M:%S'),
exchange="Binance")
list_events.append(e)
if "DEPOSIT_HISTORY_FIAT" in filename:
filename = os.path.join(folder, filename)
with open(filename, "r") as a_file:
df = pd.read_excel(filename, sheet_name="sheet1")
df = df[df["Status"] == "Succesfull"]
for index, row in df.iterrows():
e = Event(trans_type="Deposit",
buy_amt=row["Amount"],
buy_cur=row["Coin"],
fee_amt=row["Fee"],
fee_cur=row["Coin"],
comment="Binance Deposit Fiat",
effective_date=parse_date(
row["Date(UTC)"], '%Y-%m-%d %H:%M:%S'),
exchange="Binance")
list_events.append(e)
if "DEPOSIT_HISTORY_CRYPTO" in filename:
filename = os.path.join(folder, filename)
with open(filename, "r") as a_file:
df = pd.read_excel(filename, sheet_name="sheet1")
df = df[df["Status"] == "Completed"]
for index, row in df.iterrows():
e = Event(trans_type="Deposit",
buy_amt=row["Amount"],
buy_cur=row["Coin"],
comment="Binance Deposit Fiat",
effective_date=parse_date(
row["Date(UTC)"], '%Y-%m-%d %H:%M:%S'),
exchange="Binance")
list_events.append(e)
if "part" in filename:
filename = os.path.join(folder, filename)
with open(filename, "r") as a_file:
df = pd.read_csv(filename)
for index, row in df.iterrows():
row["BuyCoin"] = row["Executed"]
row["PayCoin"] = row["Amount"]
row["FeeCoin"] = row["Fee"]
row["Executed"] = row["Executed"][0:len(row["Executed"]) -
3]
row["Amount"] = row["Amount"][0:len(row["Amount"]) - 3]
row["Fee"] = row["Fee"][0:len(row["Fee"]) - 3]
row["BuyCoin"] = row["BuyCoin"].replace(
row["Executed"], "")
row["PayCoin"] = row["PayCoin"].replace(row["Amount"], "")
row["FeeCoin"] = row["FeeCoin"].replace(row["Fee"], "")
e = Event(trans_type="Trade",
buy_amt=row["Executed"],
buy_cur=row["BuyCoin"],
fee_amt=row["Fee"],
fee_cur=row["FeeCoin"],
sell_amt=row["Amount"],
sell_cur=row["PayCoin"],
comment="Binance Buy History",
effective_date=parse_date(
row["Date(UTC)"], '%Y-%m-%d %H:%M:%S'),
exchange="Binance")
list_events.append(e)
if "CONVERSIONS_BNB" in filename:
filename = os.path.join(folder, filename)
with open(filename, "r") as a_file:
df = pd.read_csv(filename)
for index, row in df.iterrows():
e = Event(trans_type="Trade",
buy_amt=row["Converted BNB"],
buy_cur="BNB",
fee_amt="BNB",
fee_cur=row["Fee (BNB)"],
sell_amt=row["Amount"],
sell_cur=row["Coin"],
comment="Binance BNB Conversion low capitals",
effective_date=parse_date(
row["Date"], '%Y-%m-%d %H:%M:%S'),
exchange="Binance")
list_events.append(e)
if "AIRDROPS" in filename:
filename = os.path.join(folder, filename)
with open(filename, "r") as a_file:
df = pd.read_csv(filename)
for index, row in df.iterrows():
e = Event(trans_type="Airdrop",
buy_amt=row["Amount"],
buy_cur=row["Coin"],
comment="Binance Airdrop " + row["Note"],
effective_date=parse_date(
row["Time"], '%Y-%m-%d %H:%M:%S'),
exchange="Binance")
list_events.append(e)
return list_events