def download_histories_csv(date_str):
filename = 'data/intraday/us.intraday.polygon.history.csv'
request_list = _get_requests(date_str)
# request_list = request_list[:10]
bt = BaseThrottler(name='base-throttler', delay=0.04)
bt.start()
throttled_requests = bt.multi_submit(request_list)
print('shutting down the throttler')
bt.shutdown()
print('waiting for the requests to be done')
bt.wait_end()
print('run_done')
responses = [tr.response for tr in throttled_requests]
with open(filename, 'w') as outfile:
outfile.write('date,time,close,open,high,low,volume,symbol\n')
for cnt, res in enumerate(responses):
if not res:
print('The response is invalid: %s' % (res))
continue
if res.status_code != 200:
continue
js = res.json()
if 'results' not in js:
print('The response does not have results: %s' % (js))
continue
data = js['results']
if not data:
continue
symbol = js['ticker']
print('{cnt}th {symbol}, blobs: {l}'.format(cnt=cnt,
symbol=symbol,
l=len(data)))
out_lines = []
for blob in data:
epoch = int(blob['t']) // 1000
t = datetime.datetime.fromtimestamp(epoch).astimezone(
_TZ_US_EAST)
date_str = t.strftime('%Y-%m-%d')
time_str = t.strftime('%H:%M:%S')
close, open_, high, low, volume = blob['c'], blob['o'], blob[
'h'], blob['l'], blob['v']
out_lines.append(
'{date_str},{time_str},{close},{open},{high},{low},{volume},{symbol}\n'
.format(date_str=date_str,
time_str=time_str,
close=close,
open=open_,
high=high,
low=low,
volume=volume,
symbol=symbol))
outfile.writelines(out_lines)
def _run_requests_return_rows(request_list):
bt = BaseThrottler(name='base-throttler', delay=0.5)
bt.start()
throttled_requests = bt.multi_submit(request_list)
print('shutting down the throttler')
bt.shutdown()
print('waiting for the requests to be done')
bt.wait_end()
print('run_done')
responses = [tr.response for tr in throttled_requests]
rows = []
for cnt, res in enumerate(responses):
if not res:
print('The response is invalid: %s' % (res))
continue
if res.status_code != 200:
continue
if not res:
print('The response does not have contents: %s' % (res))
continue
js = res.json()
if 'status' not in js or (js['status'] != 'OK'
and js['status'] != 'success'):
print('The response does not have proper status: %s' % (js))
continue
if 'tickers' not in js:
print('The response does not have results: %s' % (js))
continue
for i, ticker in enumerate(js['tickers']):
symbol = ticker['ticker']
print('{cnt}th {symbol}'.format(cnt=cnt, symbol=symbol))
daily = ticker['day']
close, open_, high, low, volume = daily['c'], daily['o'], daily[
'h'], daily['l'], daily['v']
epoch = int(ticker['1547787608999']) // 1000
t = _TZ_US_EAST.localize(datetime.datetime.fromtimestamp(epoch))
date_str = t.strftime('%Y-%m-%d')
rows.append(
'{date_str},{close},{open},{high},{low},{volume},{symbol}\n'.
format(date_str=date_str,
close=close,
open=open_,
high=high,
low=low,
volume=volume,
symbol=symbol))
return rows
def _run_requests_return_rows(request_list):
bt = BaseThrottler(name='base-throttler', delay=0.1)
bt.start()
throttled_requests = bt.multi_submit(request_list)
print('shutting down the throttler')
bt.shutdown()
print('waiting for the requests to be done')
bt.wait_end()
print('run_done')
responses = [tr.response for tr in throttled_requests]
rows = []
for cnt, response in enumerate(responses):
if not response:
print('The response is invalid: %s' % (response))
continue
if response.status_code != 200:
print('response status code is not 200 OK: {code}'.format(
code=response.status_code))
continue
if not response:
continue
js = response.json()
if not js:
print('The response is invalid: %s' % (js))
continue
if 'dataset' not in js:
print('The response does not have dataset: %s' % (js))
continue
if 'data' not in js['dataset']:
print('The response data does not have data: %s' % (js))
continue
symbol = js['dataset']['dataset_code']
data = js['dataset']['data']
for data_for_date in data:
date_str = data_for_date[0]
close, open_, high, low, volume = data_for_date[4], data_for_date[
1], data_for_date[2], data_for_date[3], data_for_date[5]
rows.append(
'{date_str},{close},{open},{high},{low},{volume},{symbol}\n'.
format(date_str=date_str,
close=close,
open=open_,
high=high,
low=low,
volume=volume,
symbol=symbol))
return rows
def main():
args = parse_args()
bt = BaseThrottler(name='base-throttler', delay=args['delay'])
reqs = []
for i in range(0, args['n_reqs']):
r = requests.Request(method='GET', url=args['url'], data='Request - ' + str(i + 1))
reqs.append(r)
with bt:
throttled_requests = bt.multi_submit(reqs)
for r in throttled_requests:
print r.response
print "Success: {s}, Failures: {f}".format(s=bt.successes, f=bt.failures)
def main():
args = parse_args()
session = requests.Session()
session.headers.update({'user-agent': 'test-user-agent'})
bt = BaseThrottler(name='base-throttler',
delay=args['delay'],
session=session)
reqs = []
for i in range(0, args['n_reqs']):
r = requests.Request(method='GET',
url=args['url'],
data='Request - ' + str(i + 1))
reqs.append(r)
with bt:
throttled_requests = bt.multi_submit(reqs)
for r in throttled_requests:
print(r.response)
print("Success: {s}, Failures: {f}".format(s=bt.successes, f=bt.failures))
class FSEconomy(object):
def __init__(self, local, user_key=None, service_key=None):
requests_cache.install_cache('fse', expire_after=3600)
self.bt = BaseThrottler(name='fse-throttler', reqs_over_time=(1, 2))
self.bt.start()
self.airports = common.load_airports()
self.aircraft = common.load_aircrafts()
self.service_key = service_key
self.user_key = user_key
if local:
self.assignments = common.load_pickled_assignments()
else:
self.assignments = self.get_assignments()
def get_aggregated_assignments(self, cargo=False):
if cargo:
self.assignments = self.assignments[self.assignments.UnitType == 'kg']
else:
self.assignments = self.assignments[self.assignments.UnitType == 'passengers']
grouped = self.assignments.groupby(['FromIcao', 'ToIcao'], as_index=False)
aggregated = grouped.aggregate(np.sum)
return aggregated.sort_values('Pay', ascending=False)
def send_single_request(self, path):
query_link = self.generate_request(const.LINK + path)
request = requests.Request(method='GET', url=query_link)
i = 0
while True:
try:
thottled_request = self.bt.submit(request)
data = thottled_request.response.content
if 'To many requests' in data or 'minimum delay' in data:
raise requests.exceptions.ConnectionError
return data
except requests.exceptions.ConnectionError:
requests_cache.clear()
if i >= 10:
raise
print 'Retrying Request'
i += 1
time.sleep(60)
def send_multi_request(self, paths):
request_queue = []
for path in paths:
query_link = self.generate_request(const.LINK + path)
request_queue.append(requests.Request(method='GET', url=query_link))
i = 0
while True:
try:
thottled_requests = self.bt.multi_submit(request_queue)
responses = [tr.response for tr in thottled_requests]
request_queue = []
complete_response = []
for response in responses:
if 'To many requests' in response.content or 'minimum delay' in response.content:
request_queue.append(response.url)
print response.content
elif 'you are now in a lockout period' in response.content:
raise Exception(response.content)
else:
complete_response.append(response.content)
if len(request_queue) > 0:
raise requests.exceptions.ConnectionError
return complete_response
except AttributeError:
for request in request_queue:
print 'Error with request: ', request
raise
except requests.exceptions.ConnectionError:
requests_cache.clear()
if i >= 10:
raise
print 'Retrying Request'
i += 1
time.sleep(60)
def get_aircrafts_by_icaos(self, icaos):
aircraft_requests = []
for icao in icaos:
aircraft_requests.append('query=icao&search=aircraft&icao={}'.format(icao))
responses = self.send_multi_request(aircraft_requests)
all_aircraft = []
for response in responses:
aircraft = pd.DataFrame.from_csv(StringIO(response))
try:
aircraft.RentalDry = aircraft.RentalDry.astype(float)
aircraft.RentalWet = aircraft.RentalWet.astype(float)
all_aircraft.append(aircraft)
except:
print 'error updating rental info: ', response
return all_aircraft
def get_assignments(self):
assignments = pd.DataFrame()
i = 0
assignment_requests = []
number_at_a_time = 1000
while i + number_at_a_time < len(self.airports):
assignment_requests.append(
'query=icao&search=jobsfrom&icaos={}'.format('-'.join(self.airports.icao[i:i + number_at_a_time])))
i += number_at_a_time
responses = self.send_multi_request(assignment_requests)
for data in responses:
assignments = pd.concat([assignments, pd.DataFrame.from_csv(StringIO(data))])
response = self.send_single_request('query=icao&search=jobsfrom&icaos={}'.format('-'.join(self.airports.icao[i:len(self.airports) - 1])))
assignments = pd.concat([assignments, pd.DataFrame.from_csv(StringIO(response))])
with open('assignments', 'wb') as f:
pickle.dump(assignments, f)
return assignments
def get_best_assignments(self, row):
df = self.assignments[(self.assignments.FromIcao == row['FromIcao']) &
(self.assignments.ToIcao == row['ToIcao']) & (self.assignments.Amount <= row['Seats'])]
if not len(df):
return None
prob = LpProblem("Knapsack problem", LpMaximize)
w_list = df.Amount.tolist()
p_list = df.Pay.tolist()
x_list = [LpVariable('x{}'.format(i), 0, 1, 'Integer') for i in range(1, 1 + len(w_list))]
prob += sum([x * p for x, p in zip(x_list, p_list)]), 'obj'
prob += sum([x * w for x, w in zip(x_list, w_list)]) <= row['Seats'], 'c1'
prob.solve()
return df.iloc[[i for i in range(len(x_list)) if x_list[i].varValue]]
def get_best_craft(self, icao, radius):
print 'Searching for the best aircraft from {}'.format(icao)
max_seats = 0
best_aircraft = None
near_icaos = self.get_closest_airports(icao, radius).icao
all_aircraft = self.get_aircrafts_by_icaos(near_icaos)
for aircraft in all_aircraft:
if not len(aircraft):
continue
merged = pd.DataFrame.merge(aircraft, self.aircraft, left_on='MakeModel', right_on='Model', how='inner')
merged = merged[
(~merged.MakeModel.isin(const.IGNORED_AIRCRAFTS)) & (merged.RentalWet + merged.RentalDry > 0)]
if not len(merged):
continue
aircraft = merged.ix[merged.Seats.idxmax()]
if aircraft.Seats > max_seats:
best_aircraft = aircraft
max_seats = aircraft.Seats
return best_aircraft
def get_closest_airports(self, icao, nm):
lat = self.airports[self.airports.icao == icao].lat.iloc[0]
nm = float(nm)
# one degree of latitude is appr. 69 nm
lat_min = lat - nm / 69
lat_max = lat + nm / 69
filtered_airports = self.airports[self.airports.lat > lat_min]
filtered_airports = filtered_airports[filtered_airports.lat < lat_max]
distance_vector = filtered_airports.icao.map(lambda x: self.get_distance(icao, x))
return filtered_airports[distance_vector < nm]
def get_distance(self, from_icao, to_icao):
try:
lat1, lon1 = [radians(x) for x in self.airports[self.airports.icao == from_icao][['lat', 'lon']].iloc[0]]
lat2, lon2 = [radians(x) for x in self.airports[self.airports.icao == to_icao][['lat', 'lon']].iloc[0]]
except IndexError:
return 9999.9
return common.get_distance(lat1, lon1, lat2, lon2)
def get_logs(self, from_id):
key = self.user_key or self.service_key
data = self.send_single_request('query=flightlogs&search=id&readaccesskey={}&fromid={}'.format(key, from_id))
logs = pd.DataFrame.from_csv(StringIO(data))
logs = logs[(logs.MakeModel != 'Airbus A321') & (logs.MakeModel != 'Boeing 737-800') & (logs.Type == 'flight')]
logs['Distance'] = logs.apply(lambda x, self=self: self.get_distance(x['From'], x['To']), axis=1)
logs = pd.merge(logs, self.aircraft, left_on='MakeModel', right_on='Model')
logs['FlightTimeH'] = logs.apply(lambda x: int(x['FlightTime'].split(':')[0]), axis=1)
logs['FlightTimeM'] = logs.apply(lambda x: int(x['FlightTime'].split(':')[1]), axis=1)
logs = logs[(logs.FlightTimeH > 0) | (logs.FlightTimeM > 0)]
logs = logs[logs.Distance > 0]
logs['AvSpeed'] = logs.apply(lambda x: 60 * x['Distance'] / (60 * x['FlightTimeH'] + x['FlightTimeM']), axis=1)
import pdb
pdb.set_trace()
def generate_request(self, query_link):
if self.user_key:
query_link += '&userkey={}'.format(self.user_key)
elif self.service_key:
query_link += '&servicekey={}'.format(self.service_key)
return query_link
def __del__(self):
self.bt.shutdown()
def _run_requests_return_rows(request_list):
bt = BaseThrottler(name='base-throttler', delay=0.04)
bt.start()
throttled_requests = bt.multi_submit(request_list)
print('shutting down the throttler')
bt.shutdown()
print('waiting for the requests to be done')
bt.wait_end()
print('run_done')
responses = [tr.response for tr in throttled_requests]
rows = []
for cnt, res in enumerate(responses):
if not res:
print('The response is invalid: %s' % (res))
continue
if res.status_code != 200:
print('response status code is not 200 OK: {code}'.format(
code=res.status_code))
continue
js = res.json()
req = request_list[cnt]
m = re.search(r'stock/([^/]+)', req.url)
if not m:
continue
if not m.groups():
continue
symbol = m.groups()[0]
if not js:
continue
print('{cnt}th {symbol}, blobs: {l}'.format(cnt=cnt,
symbol=symbol,
l=len(js)))
prev_close = None
for blob in js:
keys = ['date', 'close', 'open', 'high', 'low', 'volume']
is_blob_compromised = False
for k in keys:
if k not in blob:
print(
'blob: {blob} does not have all the expected keys, missing key: {key}'
.format(blob=str(blob), key=k))
is_blob_compromised = True
break
if is_blob_compromised:
continue
date_str = blob['date']
close, open_, high, low, volume = blob['close'], blob[
'open'], blob['high'], blob['low'], blob['volume']
if volume == '0' or volume == 0 or close is None:
close, open_, high, low = prev_close, prev_close, prev_close, prev_close
if close is None:
continue
rows.append(
'{date_str},{close},{open},{high},{low},{volume},{symbol}\n'.
format(date_str=date_str,
close=close,
open=open_,
high=high,
low=low,
volume=volume,
symbol=symbol))
prev_close = close
return rows
class LogsClient:
def __init__(self, logs_cache_dir):
self.logs_cache_dir = logs_cache_dir + '/'
self.throttler = BaseThrottler(name='base-throttler', delay=0.2)
self.throttler.start()
def fetchLogs(self, log_metadata):
log_metadata_lookup = {log[u'id']: log for log in log_metadata}
# fetch cached log data
if not os.path.isdir(self.logs_cache_dir):
os.makedirs(self.logs_cache_dir)
cache_filepaths = {
id: self.getLogFilepath(id)
for id in log_metadata_lookup
}
existing_logs = {
id: loadJson(filepath)
for id, filepath in cache_filepaths.items()
if os.path.isfile(filepath)
}
updated_log_ids = [
id for id, log in existing_logs.items()
if log[u'info'][u'date'] < log_metadata_lookup[id][u'date']
]
# fetch any new uncached logs or logs that need to be updated
fetched_log_ids = [
id for id in log_metadata_lookup
if (id not in existing_logs or id in updated_log_ids)
]
fetched_logs = {}
if len(fetched_log_ids) > 0:
reqs = [
requests.Request('GET', 'http://logs.tf/api/v1/log/' + str(id))
for id in fetched_log_ids
]
throttled_requests = self.throttler.multi_submit(reqs)
fetched_logs = {
getLogIdFromUrl(tr.request.url): tr.response.json()
for tr in throttled_requests
}
# update cache
for id, log_json in fetched_logs.items():
saveJson(self.getLogFilepath(id), log_json)
# return merged cached & new results, preferring the new results if any conflicts
return {**existing_logs, **fetched_logs}
def getUploaderLogMetadata(self, uploaderId):
return self.throttler.submit(
requests.Request(
'GET', 'http://logs.tf/api/v1/log?uploader=' + uploaderId +
'&limit=10000')).response.json()
def getLogFilepath(self, id):
return self.logs_cache_dir + str(id) + '.json'
def close(self):
self.throttler.shutdown()
def _run_requests_return_rows(request_list):
bt = BaseThrottler(name='base-throttler', delay=0.04)
bt.start()
throttled_requests = bt.multi_submit(request_list)
print('shutting down the throttler')
bt.shutdown()
print('waiting for the requests to be done')
bt.wait_end()
print('run_done')
responses = [tr.response for tr in throttled_requests]
rows = []
for cnt, res in enumerate(responses):
if not res:
print('The response is invalid: %s' % (res))
continue
if res.status_code != 200:
print('response status code is not 200 OK: {code}'.format(
code=res.status_code))
continue
if not res:
continue
js = res.json()
if 'status' not in js or (js['status'] != 'OK'
and js['status'] != 'success'):
print('The response does not have proper status: %s' % (js))
continue
keys = ['open', 'afterHours', 'high', 'low', 'volume', 'from']
is_blob_compromised = False
for k in keys:
if k not in js:
print(
'blob: {blob} does not have all the expected keys, missing key: {key}'
.format(blob=str(blob), key=k))
is_blob_compromised = True
break
if is_blob_compromised:
continue
symbol = js['symbol']
close, open_, high, low, volume = js['afterHours'], js['open'], js[
'high'], js['low'], js['volume']
print('{symbol}'.format(symbol=symbol))
close_v = float(close)
if close_v < 1.0 or close_v > 10000:
continue
date_str = datetime.datetime.strptime(
js['from'], "%Y-%m-%dT%H:%M:%SZ").strftime("%Y-%m-%d")
rows.append(
'{date_str},{close},{open},{high},{low},{volume},{symbol}\n'.
format(date_str=date_str,
close=close,
open=open_,
high=high,
low=low,
volume=volume,
symbol=symbol))
return rows