MACD hist*2 对应到各种网站上展示的结果；重新优化POC计算逻辑，和coinank上对应

chart.py

@@ -41,7 +41,7 @@ def add_ema(df, chart, period: int = 50):
 
 def add_cci(df, chart, period: int = 14, height: float = 0.1, position: str = "bottom"):
     cci = ta.CCI(df["high"], df["low"], df["close"], timeperiod=period)
-    df["cci"] = cci
+    df["CCI"] = cci
     cci_chart = chart.create_subchart(
         position=position, width=1, height=height, sync=True
     )
@@ -49,8 +49,8 @@ def add_cci(df, chart, period: int = 14, height: float = 0.1, position: str = "b
         visible=True, font_size=14, color="#FFFFFF", font_family="Times New Roman"
     )
     cci_chart.time_scale(visible=False)
-    cci_line = cci_chart.create_line(name="cci", color="#FF0000", width=2)
-    cci_line.set(df[["time", "cci"]])
+    cci_line = cci_chart.create_line(name="CCI", color="#FF0000", width=2)
+    cci_line.set(df[["time", "CCI"]])
     df = df[["time"]].copy()
     df["h"] = 100
     df["l"] = -100
@@ -74,13 +74,24 @@ def add_cci(df, chart, period: int = 14, height: float = 0.1, position: str = "b
     cci_line.set(df[["time", "l"]])
 
 
-def add_macd(df, chart, height: float = 0.1, position: str = "bottom"):
+def add_macd(
+    df,
+    chart,
+    fastperiod: int = 12,
+    slowperiod: int = 26,
+    signalperiod: int = 9,
+    height: float = 0.1,
+    position: str = "bottom",
+):
     macd, signal, hist = ta.MACD(
-        df["close"], fastperiod=12, slowperiod=26, signalperiod=9
+        df["close"],
+        fastperiod=fastperiod,
+        slowperiod=slowperiod,
+        signalperiod=signalperiod,
     )
-    df["macd"] = macd
-    df["signal"] = signal
-    df["histogram"] = hist
+    df["DIF"] = macd
+    df["DEA"] = signal
+    df["MACD"] = hist * 2
     macd_chart = chart.create_subchart(
         position=position, width=1, height=height, sync=True
     )
@@ -89,21 +100,21 @@ def add_macd(df, chart, height: float = 0.1, position: str = "bottom"):
     )
     macd_chart.time_scale(visible=False)
 
-    histogram = macd_chart.create_histogram(name="histogram")
-    hist_data = df[["time", "histogram"]].copy()
-    hist_data["color"] = hist_data["histogram"].apply(
+    histogram = macd_chart.create_histogram(name="MACD")
+    hist_data = df[["time", "MACD"]].copy()
+    hist_data["color"] = hist_data["MACD"].apply(
         lambda x: "#00FF00" if x < 0 else "#ff0000"  # 绿色 : 红色
     )
     histogram.set(hist_data)
 
     macd_line = macd_chart.create_line(
-        name="macd", color="#2962FF", width=2, price_label=False, price_line=False
+        name="DIF", color="#2962FF", width=2, price_label=False, price_line=False
     )
-    macd_line.set(df[["time", "macd"]])
+    macd_line.set(df[["time", "DIF"]])
     signal_line = macd_chart.create_line(
-        name="signal", color="#FF0000", width=2, price_label=False, price_line=False
+        name="DEA", color="#FF0000", width=2, price_label=False, price_line=False
     )
-    signal_line.set(df[["time", "signal"]])
+    signal_line.set(df[["time", "DEA"]])
 
 
 def TD(dataframe: pd.DataFrame):
@@ -260,18 +271,49 @@ def resample_data(df: pd.DataFrame, timeframe: str) -> pd.DataFrame:
 
 
 def POC(df, bins: int = 50):
+    """
+    计算价格分布的成交量峰值位置（POC，Point Of Control）
+
+    参数:
+        df: 包含 'low', 'high', 'volume' 三列的 DataFrame（每行代表一根 K 线）
+        bins: 将价格区间划分为多少个小区间（价格桶），默认为 50
+
+    返回:
+        poc: 代表成交量最大的价格区间的中点价格
+    """
+
+    # 当前数据的最低价和最高价，用于构建价格区间
     low, high = df["low"].min(), df["high"].max()
-    edges = np.linspace(low, high, bins + 1)
-    # 为每根K线生成 bins 个均匀价格点，并分配等量成交量
-    prices = np.concatenate(
-        [np.linspace(r["low"], r["high"], bins) for _, r in df.iterrows()]
-    )
-    volumes = np.repeat(df["volume"].values / bins, bins)
-    # 分箱求和
-    vol_profile, _ = np.histogram(prices, bins=edges, weights=volumes)
-    # 返回最大成交量区间的中点
-    idx = np.argmax(vol_profile)
-    poc = (edges[idx] + edges[idx + 1]) / 2
+    # 将整个价格区间等分为 bins 个小区间，需要 bins+1 个边界点
+    price_ranges = np.linspace(low, high, bins + 1)
+
+    # 初始化每个价格区间累积的成交量数组，长度为 bins（区间个数）
+    volume_per_price = np.zeros(bins)
+
+    # 遍历每一根 K 线，将该 K 线的成交量按覆盖的价格区间平均分配
+    for i in range(len(df)):
+        high = df["high"].iloc[i]
+        low = df["low"].iloc[i]
+        volume = df["volume"].iloc[i]
+
+        # 找到当前 K 线覆盖的价格边界（注意使用闭区间判断）
+        # price_ranges 表示边界点，若 price_ranges[j] 在 [low, high] 范围内，则第 j 个边界被覆盖
+        price_coverage = (price_ranges >= low) & (price_ranges <= high)
+        covered_bins = np.where(price_coverage)[0]
+
+        if len(covered_bins) > 0:
+            # 如果覆盖了多个边界点，则这些边界之间形成了若干完整的区间
+            # 将该 K 线的成交量平均分配到这些被覆盖的区间上
+            # 注意：covered_bins 里是边界索引，最后一个边界索引对应的区间索引需小于 bins
+            volume_per_bin = volume / len(covered_bins)
+            for bin_idx in covered_bins:
+                if bin_idx < bins:
+                    volume_per_price[bin_idx] += volume_per_bin
+
+    # 找到成交量最大的区间索引
+    idx = np.argmax(volume_per_price)
+    # 以该区间的两个边界的中点作为 POC 值
+    poc = (price_ranges[idx] + price_ranges[idx + 1]) / 2
     return poc
 
 
@@ -281,6 +323,7 @@ def on_range_change_poc(chart, bars_before, bars_after):
         return
 
     total_bars = len(df)
+    # TODO: k线拉到最早会报错
     if bars_after < 0:
         start_idx = max(0, int(bars_before))
         end_idx = total_bars
@@ -333,7 +376,7 @@ def plot_chart(df, symbol: str, name: str, timeframe: str):
     add_ema(df, chart, period=30)
     add_ema(df, chart, period=60)
     add_macd(df, chart)
-    add_cci(df, chart, period=14)
+    add_cci(df, chart, period=26)
     add_TD(df, chart)
     add_buy_sell_signal_markers(df, chart)
 
@@ -341,7 +384,7 @@ def plot_chart(df, symbol: str, name: str, timeframe: str):
 
 
 if __name__ == "__main__":
-    symbol = "399986"
+    symbol = "399998"
     timeframe = "1W"
 
     df = pd.read_csv(