def score(self, X, y, sample_weight=None):
p = self.predict(X)
t = y[:, 2]
pu, pl = p[:, 0], p[:, 1]
picp = picpf(t, pu, pl)
pinew = pinewf(pu, pl)
cwc = cwcf(picp, pinew, self.mu, self.eta)
return cwc
def xxx_cross_validation(t,
u,
l,
xxx_u,
xxx_l,
mu=0.9,
eta=10,
cv=3,
is_plot=False,
plt_no=None):
tscv = TimeSeriesSplit(n_splits=cv)
cwc_l = []
picp_l = []
pinew_l = []
plt_num = 0
for tr_idx, te_idx in tscv.split(t):
tr_u, te_u = u[tr_idx], u[te_idx]
tr_l, te_l = l[tr_idx], l[te_idx]
tr_t, te_t = t[tr_idx], t[te_idx]
pred_u = xxx_predict(tr_u, te_u, xxx_u, range(1, 3))
pred_l = xxx_predict(tr_l, te_l, xxx_l, range(3))
d = max(len(te_t) - len(pred_u), len(te_t) - len(pred_l))
if is_plot:
plt.figure(figsize=(8, 6))
plt.subplot(plt_no)
rd.plot_ts_ul(te_t[d:],
pred_u,
pred_l,
title='Cross validation %d' % plt_num)
plt_no += 1
plt_num += 1
picp = picpf(te_t[d:], pred_u, pred_l)
picp_l.append(picp)
pinew = pinewf(pred_u, pred_l)
pinew_l.append(pinew)
cwc = cwcf(picp, pinew, mu, eta)
cwc_l.append(cwc)
return np.array(cwc_l).mean(), np.array(picp_l).mean(), np.array(
pinew_l).mean()
def xxx_test_validation(test,
u_train,
u_test,
l_train,
l_test,
xxx_u,
xxx_l,
mu=0.9,
eta=10,
is_plot=False,
plt_num=111):
pred_u = xxx_predict(u_train, u_test, xxx_u, range(1, 3))
pred_l = xxx_predict(l_train, l_test, xxx_l, range(3))
d = max(len(test) - len(pred_u), len(test) - len(pred_l))
if is_plot:
plt.subplot(plt_num)
plt.figure(figsize=(8, 6))
rd.plot_ts_ul(test[d:], pred_u, pred_l)
picp = picpf(test[d:], pred_u, pred_l)
pinew = pinewf(pred_u, pred_l)
cwc = cwcf(picp, pinew, mu, eta)
return cwc, picp, pinew
def cross_cwcf(self,
gamma_u,
C_u,
gamma_l,
C_l,
d_u=None,
d_l=None,
cv=3,
mu=0.6,
eta=10,
isplt=False):
"""计算交叉验证的cwc（区间覆盖宽度标准） picp（区间覆盖率） pinew（区间宽度）
cv: 折数
mu: 预测置信区间
eta: 惩罚因子，表现算法偏好度。eta增加提升picp；eta减小，提升pinew
isplt: 是否将每折画图
"""
tscv = TimeSeriesSplit(n_splits=cv)
cwc_l = []
picp_l = []
pinew_l = []
i_cv = 0
plot_num = cv * 100 + 3 * 10
for tridx, teidx in tscv.split(self.cv_t):
cv_tr_u, cv_te_u = self.cv_u[tridx], self.cv_u[teidx]
cv_tr_l, cv_te_l = self.cv_l[tridx], self.cv_l[teidx]
cv_tr_t, cv_te_t = self.cv_t[tridx], self.cv_t[teidx]
svr_u = svm.SVR(gamma=gamma_u, C=C_u)
svr_l = svm.SVR(gamma=gamma_l, C=C_l)
pu, pl, d = self.svm_predict(cv_tr_u, cv_te_u, cv_tr_l, cv_te_l,
svr_u, svr_l, d_u, d_l)
if isplt:
plt.figure(figsize=(15, 6))
plt.subplot(plot_num + 1)
plt.title('cv train %d' % (i_cv))
plt.plot(cv_tr_u, color='blue', marker='o')
plt.plot(cv_tr_l, color='blue', marker='o')
plt.plot(cv_tr_t, color='gray', marker='x')
plt.subplot(plot_num + 2)
plt.title('cv test %d' % (i_cv))
plt.plot(cv_te_u, color='blue', marker='o')
plt.plot(cv_te_l, color='blue', marker='o')
plt.plot(cv_te_t, color='gray', marker='x')
plt.subplot(plot_num + 3)
plt.title('cv predict %d' % (i_cv))
plt.plot(pu, color='blue', marker='o')
plt.plot(pl, color='blue', marker='o')
plt.plot(cv_te_t[d:], color='gray', marker='x')
i_cv += 1
plot_num += 3
picp = picpf(cv_te_t[d:], pu, pl)
picp_l.append(picp)
pinew = pinewf(pu, pl)
pinew_l.append(pinew)
cwc = cwcf(picp, pinew, mu, eta)
cwc_l.append(cwc)
return np.array(cwc_l).mean(), np.array(picp_l).mean(), np.array(
pinew_l).mean()