def _check_bounds(self,x_new):
# If self.bounds_error = 1, we raise an error if any x_new values
# fall outside the range of x. Otherwise, we return an array indicating
# which values are outside the boundary region.
# !! Needs some work for multi-dimensional x !!
below_bounds = less(x_new,self.x[0])
above_bounds = greater(x_new,self.x[-1])
# Note: sometrue has been redefined to handle length 0 arrays
# !! Could provide more information about which values are out of bounds
# RHC -- Changed these ValueErrors to PyDSTool_BoundsErrors
if self.bounds_error and any(sometrue(below_bounds)):
## print "Input:", x_new
## print "Bound:", self.x[0]
## print "Difference input - bound:", x_new-self.x[0]
raise PyDSTool_BoundsError, " A value in x_new is below the"\
" interpolation range."
if self.bounds_error and any(sometrue(above_bounds)):
## print "Input:", x_new
## print "Bound:", self.x[-1]
## print "Difference input - bound:", x_new-self.x[-1]
raise PyDSTool_BoundsError, " A value in x_new is above the"\
" interpolation range."
# !! Should we emit a warning if some values are out of bounds.
# !! matlab does not.
out_of_bounds = logical_or(below_bounds,above_bounds)
return out_of_bounds
def _check_bounds(self, x_new):
# If self.bounds_error = 1, we raise an error if any x_new values
# fall outside the range of x. Otherwise, we return an array indicating
# which values are outside the boundary region.
# !! Needs some work for multi-dimensional x !!
below_bounds = less(x_new, self.x[0])
above_bounds = greater(x_new, self.x[-1])
# Note: sometrue has been redefined to handle length 0 arrays
# !! Could provide more information about which values are out of bounds
# RHC -- Changed these ValueErrors to PyDSTool_BoundsErrors
if self.bounds_error and any(sometrue(below_bounds)):
## print "Input:", x_new
## print "Bound:", self.x[0]
## print "Difference input - bound:", x_new-self.x[0]
raise PyDSTool_BoundsError, " A value in x_new is below the"\
" interpolation range."
if self.bounds_error and any(sometrue(above_bounds)):
## print "Input:", x_new
## print "Bound:", self.x[-1]
## print "Difference input - bound:", x_new-self.x[-1]
raise PyDSTool_BoundsError, " A value in x_new is above the"\
" interpolation range."
# !! Should we emit a warning if some values are out of bounds.
# !! matlab does not.
out_of_bounds = logical_or(below_bounds, above_bounds)
return out_of_bounds
def compute(self, frame, lamda):
Y = sp.fft(frame * self._window)
Ysq = sp.absolute(Y)**2.0
#Compute smooth speech power spectrum
P = self._nu * self._lastP + (1.0 - self._nu) * Ysq #eq 2
#Find the local minimum of noisy speech
Pmin = sp.zeros(self._winsize, dtype=sp.float32)
minidx = sp.less_equal(P, self._lastPmin)
largeidx = sp.greater(P, self._lastPmin)
Pmin[minidx] = P[minidx]
Pmin[largeidx]=self._gamma * self._lastPmin[largeidx] + \
((1.0-self._gamma)/(1.0-self._B)) * (P[largeidx] - self._B* self._lastP[largeidx])#eq3
#Compute ratio of smoothed speech p
Sr = P / Pmin #eq4
#Calculate speech presence probability using first-order recursion
I = sp.array(sp.greater(self._delta, Sr), dtype=sp.float32) #eq5
p = self._alpha_p * self._lastp + (1.0 - self._alpha_p) * I #eq6
#Compute time-frequency dependent smoothing factors
alpha_s = self._alpha_d + (1.0 - self._alpha_d) * p #eq7
#Update noise estimate using time-frequency dependent smoothing factors
D = alpha_s * self._lastD + (1.0 - alpha_s) * Ysq #eq8
#calculate debug info
noise_periodgram = 10 * sp.log10(
sp.sum(sp.absolute(D)) / self._winsize)
noisy_speech_periodgram = 10 * sp.log10(sp.sum(Ysq) / self._winsize)
smooth_noisy_speech_periodgram = 10 * sp.log10(
sp.sum(P) / self._winsize)
self._nplist.append(noise_periodgram)
self._nsplist.append(noisy_speech_periodgram)
self._snsplist.append(smooth_noisy_speech_periodgram)
#update
self._lastP = P
self._lastPmin = Pmin
self._lastp = p
self._lastD = D
return D
def compute(self,frame,lamda):
Y = sp.fft(frame*self._window)
Ysq = sp.absolute(Y)**2.0
#Compute smooth speech power spectrum
P=self._nu*self._lastP + (1.0-self._nu)*Ysq#eq 2
#Find the local minimum of noisy speech
Pmin = sp.zeros(self._winsize,dtype=sp.float32)
minidx = sp.less_equal(P,self._lastPmin)
largeidx = sp.greater(P,self._lastPmin)
Pmin[minidx]=P[minidx]
Pmin[largeidx]=self._gamma * self._lastPmin[largeidx] + \
((1.0-self._gamma)/(1.0-self._B)) * (P[largeidx] - self._B* self._lastP[largeidx])#eq3
#Compute ratio of smoothed speech p
Sr = P/Pmin #eq4
#Calculate speech presence probability using first-order recursion
I = sp.array(sp.greater(self._delta,Sr),dtype=sp.float32)#eq5
p = self._alpha_p * self._lastp + (1.0 - self._alpha_p)*I#eq6
#Compute time-frequency dependent smoothing factors
alpha_s = self._alpha_d + (1.0 - self._alpha_d)*p#eq7
#Update noise estimate using time-frequency dependent smoothing factors
D = alpha_s*self._lastD + (1.0 - alpha_s)*Ysq#eq8
#calculate debug info
noise_periodgram = 10*sp.log10(sp.sum(sp.absolute(D))/self._winsize)
noisy_speech_periodgram = 10*sp.log10(sp.sum(Ysq)/self._winsize)
smooth_noisy_speech_periodgram = 10*sp.log10(sp.sum(P)/self._winsize)
self._nplist.append(noise_periodgram)
self._nsplist.append(noisy_speech_periodgram)
self._snsplist.append(smooth_noisy_speech_periodgram)
#update
self._lastP = P
self._lastPmin = Pmin
self._lastp = p
self._lastD = D
return D
def update_stats(self):
self.accepted=sp.sum(self.chain_accept)
self.sample_size=sp.sum(self.chain_sizes)
#Update means and errors if any chain contains acceptable points
#print 'pre state:',self.mean
if float(sp.sum(self.chain_weights))>0:
#Total mean
#for i in range(len(self.chain_weights)):
#print self.chain_weights[i]*self.chain_means[i]
self.mean=sp.sum([self.chain_weights[i]*self.chain_means[i] for i in range(len(self.chain_weights)) if not sp.isnan(self.chain_means[i]).all()],0)/float(sp.sum(self.chain_weights))
#Total error (assumes no correlation at present)
if sp.greater(self.chain_sizes,self.batch_size).any():
w=sp.array(self.chain_weights)/float(sp.sum(self.chain_weights))
self.mean_error=sp.sqrt(sp.sum([w[i]**2*self.chain_mean_errors[i]**2 for i in range(len(w)) if self.chain_sizes[i]>self.batch_size],0))
def stats(var, tmpFilters, tmpReturns):
results = {'total': None, 'win_ct': None, 'lose_ct': None, 'win_ratio': None,
'lose_ratio': None, 'return_med': None, 'return_avg': None, 'return_stddev': None,
'return_min': None, 'return_max': None, 'slope': None, 'intercept': None, 'r': None, 'r_low': None,
'r_high': None, '2_tail_prob': None, 'std_err': None}
total = float( len(tmpReturns) )
winct = float( sp.greater( tmpReturns, 0 ).sum() )
losect = total - winct
if total > 0:
winrt = winct / total
losert = losect / total
else:
winrt = 0
losert = 0
if len(tmpReturns) > 0:
returnMed = sp.median( tmpReturns )
returnAvg = sp.mean( tmpReturns )
returnStdDev = sp.std( tmpReturns )
returnMin = np.min( tmpReturns )
returnMax = np.max( tmpReturns )
else:
returnMed = 0
returnAvg = 0
returnStdDev = 0
returnMin = 0
returnMax = 0
if total > 0 and var != None:
r = scipy.stats.linregress( tmpFilters, tmpReturns )
corr = r[2]
z_r = np.arctanh(corr)
ci = 1.96
z_low = z_r - ci/np.sqrt(len(tmpReturns)-3)
z_high = z_r + ci/np.sqrt(len(tmpReturns)-3)
r_low = ( np.exp(1) ** ( 2 * z_low ) - 1 ) / ( np.exp(1) ** ( 2 * z_low ) + 1 )
r_high = ( np.exp(1) ** ( 2 * z_high ) - 1 ) / ( np.exp(1) ** ( 2 * z_high ) + 1 )
slope = r[0]
intercept = r[1]
twoTail = r[3]
stdErr = r[4]
else:
corr = 0
r_low = 0
r_high = 0
slope = 0
intercept = 0
twoTail = 0
stdErr = 0
if len(tmpReturns) > 0:
results = { 'total': total,
'win_ct': winct,
'lose_ct': losect,
'win_ratio': winrt,
'lose_ratio': losert,
'return_med': returnMed,
'return_avg': returnAvg,
'return_stddev': returnStdDev,
'return_min': returnMin,
'return_max': returnMax,
'slope': slope,
'intercept': intercept,
'r': corr,
'r_low': r_low,
'r_high': r_high,
'2_tail_prob': twoTail,
'std_err': stdErr}
return results
def run_simulation_fast(vol, lam, sprd_client, sprd_dealer, delta_lim,
hedge_style, dt, nsteps, nruns, seed):
'''Runs a Monte Carlo simulation and returns statics on PNL, client trades, and hedge trades.
"_fast" because it uses vectorized operations.
vol: lognormal volatility of the spot process
lam: Poisson process frequency
sprd_client: fractional bid/ask spread for client trades. eg 1e-4 means 1bp.
sprd_dealer: fractional bid/ask spread for inter-dealer hedge trades. eg 1e-4 means 1bp.
delta_lim: the delta limit at or beyond which the machine will hedge in the inter-dealer market
hedge_style: 'Zero' or 'Edge', defining the hedging style. 'Zero' means hedge to zero position,
'Edge' means hedge to the nearer delta limit.
dt: length of a time step
nsteps: number of time steps for each run of the simulation
nruns: number of Monte Carlo runs
seed: RNG seed
'''
scipy.random.seed(seed)
trade_prob = 1 - exp(-lam * dt)
sqrtdt = sqrt(dt)
spots = scipy.zeros(nruns) + 1 # initial spot == 1
posns = scipy.zeros(nruns)
trades = scipy.zeros(nruns)
hedges = scipy.zeros(nruns)
pnls = scipy.zeros(nruns)
for step in range(nsteps):
dzs = scipy.random.normal(0, sqrtdt, nruns)
qs = scipy.random.uniform(0, 1, nruns)
ps = scipy.random.binomial(
1, 0.5,
nruns) * 2 - 1 # +1 or -1 - trade quantities if a trade happens
# check if there are client trades for each path
indics = scipy.less(qs, trade_prob)
posns += indics * ps
trades += scipy.ones(nruns) * indics
pnls += scipy.ones(nruns) * indics * sprd_client * spots / 2.
# check if there are hedges to do for each path
if hedge_style == 'Zero':
indics = scipy.logical_or(scipy.less_equal(posns, -delta_lim),
scipy.greater_equal(posns, delta_lim))
pnls -= scipy.absolute(posns) * indics * sprd_dealer * spots / 2.
posns -= posns * indics
hedges += scipy.ones(nruns) * indics
elif hedge_style == 'Edge':
# first deal with cases where pos>delta_lim
indics = scipy.greater(posns, delta_lim)
pnls -= (posns - delta_lim) * indics * sprd_dealer * spots / 2.
posns = posns * scipy.logical_not(indics) + scipy.ones(
nruns) * indics * delta_lim
hedges += scipy.ones(nruns) * indics
# then the cases where pos<-delta_lim
indics = scipy.less(posns, -delta_lim)
pnls -= (-delta_lim - posns) * indics * sprd_dealer * spots / 2.
posns = posns * scipy.logical_not(indics) + scipy.ones(
nruns) * indics * (-delta_lim)
hedges += scipy.ones(nruns) * indics
else:
raise ValueError('hedge_style must be "Edge" or "Zero"')
# advance the spots and calculate period PNL
dspots = vol * spots * dzs
pnls += posns * dspots
spots += dspots
return {
'PNL': (pnls.mean(), pnls.std()),
'Trades': (trades.mean(), trades.std()),
'Hedges': (hedges.mean(), hedges.std())
}
