如何将Pinescript中zoneDecider相关代码转换为Python代码？

已完成的Python代码 #

df = data
    
zone_length = 10
zone_multiplier = 3.0
df['hl2'] = (df['h'] + df['l']) / 2
df.ta.atr(high=df['h'], low=df['l'], close=df['c'], length=zone_length, append=True)
df['downZone'] = df['hl2'] + (zone_multiplier * df['ATRr_10'])
df['upZone'] = df['hl2'] - (zone_multiplier * df['ATRr_10'])

待转换的PineScript代码 #

downZone = hl2 + (3.0 * ta.atr(10))
upZone = hl2 - (3.0 * ta.atr(10) )

downZoneNew = nz(downZone[1], downZone)
downZone := close[1] < downZoneNew ? math.min(downZone, downZoneNew) : downZone

upZoneNew = nz(upZone[1], upZone)
upZone := close[1] > upZoneNew ? math.max(upZone, upZoneNew) : upZone

zoneDecider = 1
zoneDecider := nz(zoneDecider[1], zoneDecider)
zoneDecider := zoneDecider == -1 and close > downZoneNew ? 1 : zoneDecider == 1 and close < upZoneNew ? -1 : zoneDecider

redZone = zoneDecider == -1 and zoneDecider[1] == 1
greenZone = zoneDecider == 1 and zoneDecider[1] == -1


plotshape(redZone ? downZone : na, location=location.absolute, style=shape.labeldown, size=size.tiny, color=color.new(color.red, 0), textcolor=color.new(color.white, 0), title='SAT', text='SAT')
plotshape(greenZone ? upZone : na, location=location.belowbar, style=shape.labelup, size=size.tiny, color=color.new(color.green, 0), textcolor=color.new(color.white, 0), title='AL', text='AL') 

转换后的Python实现 #

由于PineScript是逐K线迭代计算，且变量更新依赖前一根K线的结果，Python中需要用循环模拟该逻辑：

import pandas as pd

# 初始化迭代用的前值变量
down_zone_prev = df['downZone'].iloc[0]
up_zone_prev = df['upZone'].iloc[0]
zone_decider_prev = 1

# 创建新列存储最终计算结果
df['downZone_final'] = df['downZone']
df['upZone_final'] = df['upZone']
df['zoneDecider'] = 1
df['redZone'] = False
df['greenZone'] = False

# 从第二根K线开始逐行计算
for i in range(1, len(df)):
    close_prev = df['c'].iloc[i-1]
    current_downZone = df['downZone'].iloc[i]
    current_upZone = df['upZone'].iloc[i]
    
    # 处理downZone更新逻辑
    downZoneNew = down_zone_prev if not pd.isna(down_zone_prev) else current_downZone
    updated_downZone = min(current_downZone, downZoneNew) if close_prev < downZoneNew else current_downZone
    df.at[i, 'downZone_final'] = updated_downZone
    
    # 处理upZone更新逻辑
    upZoneNew = up_zone_prev if not pd.isna(up_zone_prev) else current_upZone
    updated_upZone = max(current_upZone, upZoneNew) if close_prev > upZoneNew else current_upZone
    df.at[i, 'upZone_final'] = updated_upZone
    
    # 更新zoneDecider状态
    zone_decider_current = zone_decider_prev
    if zone_decider_current == -1 and df['c'].iloc[i] > downZoneNew:
        zone_decider_current = 1
    elif zone_decider_current == 1 and df['c'].iloc[i] < upZoneNew:
        zone_decider_current = -1
    df.at[i, 'zoneDecider'] = zone_decider_current
    
    # 识别状态反转的zone信号
    df.at[i, 'redZone'] = (zone_decider_current == -1) and (zone_decider_prev == 1)
    df.at[i, 'greenZone'] = (zone_decider_current == 1) and (zone_decider_prev == -1)
    
    # 更新前值，用于下一轮迭代
    down_zone_prev = updated_downZone
    up_zone_prev = updated_upZone
    zone_decider_prev = zone_decider_current

关键转换说明 #

• nz()函数替代：用pd.isna()判断空值，取前一根K线的结果或当前值作为默认值。
• 逐行迭代逻辑：PineScript的:=是逐K线更新变量，Python中无法用向量操作直接实现，必须通过循环逐行计算并传递前值。
• 状态反转判断：直接对比当前和前一根K线的zoneDecider值，得到redZone和greenZone信号。
• 绘图替代：如果需要实现类似PineScript的标注效果，可以用matplotlib或plotly库，在对应K线的价格位置添加文本标签。

内容的提问来源于stack exchange，提问作者abaci