def history_to_string(self, last=True):
"""
Generates a string representing the recorded history of an order
:return:
"""
s = ''
if last:
if self.type in ['A', 'F', 'U']:
lasttype = self.history[-1][
0] # Type of the last order applied to the order
if lasttype == 'C':
s += 'Last: ' + lasttype + ' DeltaT: ' + nanoseconds_to_time(self.history[-1][1] - self.otime) +\
' SZ:' + str(self.history[-1][2]) + ' P:' + str(self.history[-1][3])
if lasttype == 'E':
delta = self.history[-1][1] - self.otime
s += f'Last: {lasttype}' + \
f' DeltaT: {nanoseconds_to_time(delta)}' + \
f' SZ: {self.history[-1][2]}'
if delta < 1_000_000:
s = '**' + s
if delta < 1_000_000_000:
s = '*' + s
h = 'HIST = '
for o in self.history:
h += f'{o[0]}->'
s += f'\n{h}'
if lasttype == 'X':
s += 'Last: ' + lasttype + ' DeltaT: ' + nanoseconds_to_time(self.history[-1][1] - self.otime) +\
' SZ:' + str(self.history[-1][2])
if lasttype == 'D':
s += 'Last: ' + lasttype + ' DeltaT: ' + nanoseconds_to_time(
self.history[-1][1] - self.otime)
if lasttype == 'U':
s += 'H: ' + lasttype + ' DeltaT: ' + nanoseconds_to_time(self.history[-1][1] - self.otime) +\
' SZ:' + str(self.history[-1][2]) + ' P:' + str(self.history[-1][3])
return s
def process_order(self, order):
"""
Updates the counter
:param order:
:return:
"""
if order.type in self.selected:
if self.counter[order.type][0] == int(order.otime/self.granularity):
self.counter[order.type][1] += 1
else:
self.countertime[order.type].append(self.counter[order.type][0])
self.countervalue[order.type].append(self.counter[order.type][1])
self.countertick[order.type].append(nanoseconds_to_time(self.counter[order.type][0]*self.granularity, prec=self.precision))
self.counter[order.type] = [int(order.otime/self.granularity), 1]
if self.counter['Z'][0] == int(order.otime/self.granularity):
self.counter['Z'][1] += 1
else:
self.countertime['Z'].append(self.counter['Z'][0])
self.countervalue['Z'].append(self.counter['Z'][1])
self.countertick['Z'].append(nanoseconds_to_time(self.counter['Z'][0]*self.granularity, prec=self.precision))
self.counter['Z'] = [int(order.otime/self.granularity), 1]
def log_process(statistics, log=False):
"""
Some logging of the results
:param statistics:
:return:
"""
if log:
print('N Buy orders:', len(statistics['buy']['ordertime']))
print('N Sell orders:', len(statistics['sell']['ordertime']))
print('N Order Executions Sell:',
len(statistics['sell']['executiondeltatime']))
print(
'Mean time to execution:',
nanoseconds_to_time(
np.mean(statistics['sell']['executiondeltatime'])))
print(
'Max time to execution:',
nanoseconds_to_time(
np.max(statistics['sell']['executiondeltatime'])))
print(
'Min time to execution:',
nanoseconds_to_time(
np.min(statistics['sell']['executiondeltatime'])))
print('N Order Executions Buy:',
len(statistics['buy']['executiondeltatime']))
print(
'Mean time to execution:',
nanoseconds_to_time(
np.mean(statistics['buy']['executiondeltatime'])))
print(
'Max time to execution:',
nanoseconds_to_time(np.max(
statistics['buy']['executiondeltatime'])))
print(
'Min time to execution:',
nanoseconds_to_time(np.min(
statistics['buy']['executiondeltatime'])))
def log(self):
"""
Data info
:return:
"""
if self.day is None:
raise NameError('Montly data are not loaded')
print('N Buy orders:', len(self.statistics['buy']['ordertime']))
print('N Sell orders:', len(self.statistics['sell']['ordertime']))
print('N Order Executions Sell:',
len(self.statistics['sell']['executiondeltatime']))
print(
'Mean time to execution:',
nanoseconds_to_time(
np.mean(self.statistics['sell']['executiondeltatime'])))
print(
'Max time to execution:',
nanoseconds_to_time(
np.max(self.statistics['sell']['executiondeltatime'])))
print(
'Min time to execution:',
nanoseconds_to_time(
np.min(self.statistics['sell']['executiondeltatime'])))
print('N Order Executions Buy:',
len(self.statistics['buy']['executiondeltatime']))
print(
'Mean time to execution:',
nanoseconds_to_time(
np.mean(self.statistics['buy']['executiondeltatime'])))
print(
'Max time to execution:',
nanoseconds_to_time(
np.max(self.statistics['buy']['executiondeltatime'])))
print(
'Min time to execution:',
nanoseconds_to_time(
np.min(self.statistics['buy']['executiondeltatime'])))
ltimeEB.append(order.otime)
lpriceEB.append(trans.price)
lsizeEB.append(trans.size)
# Non-displayable orders
if order.type in ['P']:
ltimeEP.append(order.otime)
lpriceEP.append(order.price)
lsizeEP.append(order.size)
print('Stock:', stock, 'Day:', day)
print('N Buy orders:', len(ltimeOB))
print('N Sell orders:', len(ltimeOS))
print('N Order Executions Sell:', len(lexecutionsS))
print('Mean time to execution:',
nanoseconds_to_time(np.mean(lexecutionsS)))
print('Max time to execution:',
nanoseconds_to_time(np.max(lexecutionsS)))
print('Min time to execution:',
nanoseconds_to_time(np.min(lexecutionsS)))
print('N Order Executions Buy:', len(lexecutionsB))
print('Mean time to execution:',
nanoseconds_to_time(np.mean(lexecutionsB)))
print('Max time to execution:',
nanoseconds_to_time(np.max(lexecutionsB)))
print('Min time to execution:',
nanoseconds_to_time(np.min(lexecutionsB)))
print('N Hiden Executions:', len(ltimeEP))
ddata = {}
lpriceEP.append(float(data[7].strip()))
lsizeEP.append(int(data[6].strip()))
i += 1
if i % 10000 == 0:
print('.', end='', flush=True)
print()
print('Stock:', stock)
if cpny.get_company(stock) is not None:
print(cpny.get_company(stock))
print('N Buy/Sell orders:', norders)
if len(lexecutionsS) != 0:
print('N Order Executions Sell:', len(lexecutionsS))
print('Mean time to execution:',
nanoseconds_to_time(np.mean(lexecutionsS)))
print('Max time to execution:',
nanoseconds_to_time(np.max(lexecutionsS)))
print('Min time to execution:',
nanoseconds_to_time(np.min(lexecutionsS)))
ax = sns.distplot(np.log10(lexecutionsS), kde=True, norm_hist=True)
plt.title('Log plot of Sell execution time ' + day)
plt.show()
plt.close()
ax = sns.distplot(capped_prices(lpriceOS), kde=True, norm_hist=True)
plt.title('Orders Sell price ' + day)
plt.show()
plt.close()
if len(lexecutionsB) != 0:
print('N Order Executions Buy:', len(lexecutionsB))