實施提議方式的實現(xiàn)

和往常一樣，理論上一切看起來都很完美。 現(xiàn)在，我們來檢查一下從兩個分類器中獲得的實際效果。 為此，我們需要再次重寫代碼。

成交自動標(biāo)記的函數(shù)

我已做了一些改動。 現(xiàn)在可以根據(jù)元模型標(biāo)簽重新標(biāo)記基準(zhǔn)模型的標(biāo)簽：

def labelling_relabeling(dataset, min=15, max=15, relabeling=False) -> pd.DataFrame: ????labels = [] ????for i in range(dataset.shape[0]-max): ????????rand = random.randint(min, max) ????????curr_pr = dataset['close'][i] ????????future_pr = dataset['close'][i + rand] ????????if relabeling:????????????m_labels??= dataset['meta_labels'][i:rand+1].values???????? ????????if relabeling and 0.0 in m_labels:????????????labels.append(2.0)????????else: ????????????if future_pr + MARKUP < curr_pr: ????????????????labels.append(1.0) ????????????elif future_pr - MARKUP > curr_pr: ????????????????labels.append(0.0) ????????????else: ????????????????labels.append(2.0) ???????? ????dataset = dataset.iloc[:len(labels)].copy() ????dataset['labels'] = labels ????dataset = dataset.dropna() ????dataset = dataset.drop( ????????dataset[dataset.labels == 2].index) ????return dataset

高亮顯示的代碼檢查是否存在重新標(biāo)記標(biāo)志。 如果它為真，并且當(dāng)前交易水平線元標(biāo)記包含零，則元模型拒絕在該部分進(jìn)行交易。 相應(yīng)地，此類成交被標(biāo)記為 2.0，并從數(shù)據(jù)集合中刪除。 因此，我們能夠從基準(zhǔn)模型的訓(xùn)練樣本中迭代去除不必要的樣本，從而減少其訓(xùn)練的誤差。

自定義測試器函數(shù)

現(xiàn)在有一個擴(kuò)展功能，允許我們同時測試兩個模型（基準(zhǔn)模型和元模型）。 此外，定制測試器現(xiàn)在能夠重新標(biāo)記元模型的標(biāo)簽，以便在下一次迭代中改進(jìn)它。

def tester(dataset: pd.DataFrame, markup=0.0, use_meta=False, plot=False): ????last_deal = int(2) ????last_price = 0.0????report = [0.0] ????meta_labels = dataset['labels'].copy() ????for i in range(dataset.shape[0]): ????????pred = dataset['labels'][i] ????????meta_labels[i] = np.nan ????????if use_meta:????????????pred_meta = dataset['meta_labels'][i] # 1 = allow trades????????if last_deal == 2 and ((use_meta and pred_meta==1) or not use_meta): ????????????last_price = dataset['close'][i] ????????????last_deal = 0 if pred <= 0.5 else 1????????????continue????????if last_deal == 0 and pred > 0.5 and ((use_meta and pred_meta==1) or not use_meta): ????????????last_deal = 2????????????report.append(report[-1] - markup + ??????????????????????????(dataset['close'][i] - last_price)) ????????????if report[-1] > report[-2]:????????????????meta_labels[i] = 1????????????else: ????????????????meta_labels[i] = 0????????????continue????????if last_deal == 1 and pred < 0.5 and ((use_meta and pred_meta==1) or not use_meta): ????????????last_deal = 2????????????report.append(report[-1] - markup + ??????????????????????????(last_price - dataset['close'][i])) ????????????if report[-1] > report[-2]:????????????????meta_labels[i] = 1????????????else: ????????????????meta_labels[i] = 0????y = np.array(report).reshape(-1, 1) ????X = np.arange(len(report)).reshape(-1, 1) ????lr = LinearRegression() ????lr.fit(X, y) ????l = lr.coef_ ????if l >= 0: ????????l = 1????else: ????????l = -1????if(plot): ????????plt.plot(report) ????????plt.plot(lr.predict(X)) ????????plt.title("Strategy performance R^2 " + str(format(lr.score(X, y) * l,".2f"))) ????????plt.xlabel("the number of trades") ????????plt.ylabel("cumulative profit in pips") ????????plt.show() ????return lr.score(X, y) * l, meta_labels.fillna(method='backfill')