Skip to content

Commit

Permalink
.
Browse files Browse the repository at this point in the history
  • Loading branch information
@kanghua309
kanghua309 committed Dec 23, 2017
1 parent 34f327e commit 6a49914
Showing 1 changed file with 32 additions and 36 deletions.
68 changes: 32 additions & 36 deletions campaign/riskmanager/strategy_with_portfolio_optim.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -87,26 +87,26 @@



def Markowitz(inputs, mask ):
def Markowitz(inputs, window_length,mask ):
class Markowitz(CustomFactor):
#params = {'trigger_date': None, }
window_length = int(1)
target_ret = 0.01 # TODO
window_length = int(5)
target_ret = 0.1 # TODO
max_sector_exposure = 0.1
def compute(self, today, assets,out,returns,*factors):
#print("------------------------------- Markowitz:",today)
#print ("Markowitz factor:",today)
def compute(self, today, assets,out,returns):
print("------------------------------- Markowitz:",today)
#print ("Markowitz factor:",returns)
gamma = cvx.Parameter(sign="positive")
gamma.value = 1 # gamma is a Parameter that trades off riskmanager and return.
returns = np.nan_to_num(returns.T) # time,stock to stock,time
# [[1 3 2] [3 2 1]] = > [[1 3] [3 2] [2 1]]
#print ("Markowitz return ...\n", returns)
cov_mat = np.cov(returns)
Sigma = cov_mat

#returns = np.nan_to_num(returns.T) # time,stock to stock,time
#print ("Markowitz factor2:",returns)
#cov_mat = np.cov(returns)
#Sigma = cov_mat
Sigma = np.nan_to_num(returns)
Sigma = Sigma.T.dot(Sigma)
#print ("Markowitz factor2:",Sigma)
########################################################
w = cvx.Variable(len(assets))
f = F.T * w
risk = cvx.quad_form(w, Sigma)
mu = np.array([self.target_ret] * len(assets))
expected_return = np.reshape(mu,(-1, 1)).T * w # w is a vector of stock holdings as fractions of total assets.
Expand All @@ -124,12 +124,12 @@ def compute(self, today, assets,out,returns,*factors):
if classid not in sector_dist:
_ = []
sector_dist[classid] = _
sector_dist[classid].append(idx)
sector_dist[classid].append(idx)
idx += 1
#print"sector size :", len(sector_dist)
#print("sector size :", len(sector_dist),idx,sector_dist)
for k, v in sector_dist.items():
constraints.append(cvx.sum_entries(w[v]) < (1 + self.max_sector_exposure) / len(sector_dist))
constraints.append(cvx.sum_entries(w[v]) >= (1 - self.max_sector_exposure) / len(sector_dist))
constraints.append(cvx.sum_entries(w[v]) < (1 / len(sector_dist) + self.max_sector_exposure))
constraints.append(cvx.sum_entries(w[v]) >= (1 / len(sector_dist) - self.max_sector_exposure))

prob = cvx.Problem(objective, constraints)
prob.solve()
Expand All @@ -138,7 +138,7 @@ def compute(self, today, assets,out,returns,*factors):
return None
#print ("Markowit weight",np.squeeze(np.asarray(w.value))) # Remo
out[:] = np.squeeze(np.asarray(w.value)) #每行中的列1
return Markowitz(inputs = inputs,mask = mask)
return Markowitz(inputs = inputs , window_length=window_length , mask = mask)


def BasicFactorRegress(inputs, window_length, mask, n_fwd_days, algo_mode=None, cross=True):
Expand Down Expand Up @@ -203,7 +203,7 @@ def make_pipeline(asset_finder, algo_mode):
private_universe = private_universe_mask(hs300.tolist(), asset_finder=asset_finder)
#private_universe = private_universe_mask( ['000001','000002','000005'],asset_finder=asset_finder)
######################################################################################################
returns = Returns(inputs=[USEquityPricing.close], window_length=5, mask=private_universe).downsample('week_start') # 预测一周数据
returns = Returns(inputs=[USEquityPricing.close], window_length=5, mask=private_universe) # 预测一周数据
######################################################################################################
pe = Fundamental(mask=private_universe, asset_finder=asset_finder).pe.downsample('month_start')
pb = Fundamental(mask=private_universe, asset_finder=asset_finder).pb.downsample('month_start')
Expand All @@ -218,21 +218,21 @@ def make_pipeline(asset_finder, algo_mode):
gpr = Fundamental(mask=private_universe, asset_finder=asset_finder).gpr.downsample('month_start')
npr = Fundamental(mask=private_universe, asset_finder=asset_finder).npr.downsample('month_start')

rev10 = Returns(inputs=[USEquityPricing.close], window_length=10, mask=private_universe).downsample('week_start')
vol10 = AverageDollarVolume(window_length=20, mask=private_universe).downsample('week_start')
rev20 = Returns(inputs=[USEquityPricing.close], window_length=20, mask=private_universe).downsample('week_start')
vol20 = AverageDollarVolume(window_length=20, mask=private_universe).downsample('week_start')
rev30 = Returns(inputs=[USEquityPricing.close], window_length=30, mask=private_universe).downsample('week_start')
vol30 = AverageDollarVolume(window_length=20, mask=private_universe).downsample('week_start')
rev10 = Returns(inputs=[USEquityPricing.close], window_length=10, mask=private_universe)
vol10 = AverageDollarVolume(window_length=20, mask=private_universe)
rev20 = Returns(inputs=[USEquityPricing.close], window_length=20, mask=private_universe)
vol20 = AverageDollarVolume(window_length=20, mask=private_universe)
rev30 = Returns(inputs=[USEquityPricing.close], window_length=30, mask=private_universe)
vol30 = AverageDollarVolume(window_length=20, mask=private_universe)

illiq22 = ILLIQ(window_length=22, mask=private_universe).downsample('week_start')
illiq5 = ILLIQ(window_length=5, mask=private_universe).downsample('week_start')
illiq22 = ILLIQ(window_length=22, mask=private_universe)
illiq5 = ILLIQ(window_length=5, mask=private_universe)

rsi5 = RSI(window_length=5, mask=private_universe).downsample('week_start')
rsi22 = RSI(window_length=22, mask=private_universe).downsample('week_start')
rsi5 = RSI(window_length=5, mask=private_universe)
rsi22 = RSI(window_length=22, mask=private_universe)

mom5 = Momentum(window_length=5, mask=private_universe).downsample('week_start')
mom22 = Momentum(window_length=22, mask=private_universe).downsample('week_start')
mom5 = Momentum(window_length=5, mask=private_universe)
mom22 = Momentum(window_length=22, mask=private_universe)

sector = get_sector(asset_finder=asset_finder, mask=private_universe).downsample('week_start')
ONEHOTCLASS, sector_indict_keys = get_sector_by_onehot(asset_finder=asset_finder, mask=private_universe)
Expand Down Expand Up @@ -299,13 +299,9 @@ def make_pipeline(asset_finder, algo_mode):
long_short_screen = (longs | shorts)


weights = Markowitz(inputs=factors_pipe.values(),mask=long_short_screen).downsample('week_start')
weights = Markowitz(inputs=[returns],window_length=21,mask=long_short_screen).downsample('week_start')
# TODO sector onehot
pipe_final_columns = {
'Predict Factor': predict,
#'longs': longs,
#'shorts': shorts,
'predict_rank': predict_rank,
'weights': weights,
}

Expand Down

0 comments on commit 6a49914

Please sign in to comment.