Pine Script v5多时间多条件plotshape()失效问题求助

在Pine Script v5中使用plotshape()时遇到以下异常：

• 结合2分钟周期自身条件与通过request.security()获取的5分钟周期条件绘制信号
• 仅加入单个跨周期条件时，plotshape()能正常显示信号
• 同时加入2个及以上跨周期条件时，plotshape()完全不显示信号，但跨周期数据hist5Min的历史值可正常绘制

需要排查是代码错误还是plotshape()函数的bug，原代码如下：

//////// Long and Short Signals by MACD Histogram, 
//////// 2 minutes(indicator) AND 5 minutes(through request.security()

indicator(title="Moving Average Convergence Divergence Cross w/ Fill & Signal", shorttitle="MACD Cross w/ Fill & Signal")

// Calculate 2 Min MACD
fast_ma = fast_ma_type == "SMA" ? ta.sma(src, fast_len) : ta.ema(src, fast_len)
slow_ma = fast_ma_type == "SMA" ? ta.sma(src, slow_len) : ta.ema(src, slow_len)
macd = fast_ma - slow_ma
signal = slow_ma_type == "SMA" ? ta.sma(macd, signal_len) : ta.ema(macd, signal_len)
hist = macd - signal

// Get 5 Min Histogram Data
[_, _, hist5] = ta.macd(src, fast_len, slow_len, signal_len)
hist5Min = request.security(syminfo.tickerid, "5", hist5, gaps = barmerge.gaps_off)

// Go Long conditions
go_long2 = hist > hist[1]
go_long5A = hist5Min > hist5Min[1]
go_long5B = hist5Min[1] < hist5Min[2]

// Go Short conditions
go_short2 = hist < hist[1]
go_short5A = hist5Min < hist5Min[1]
go_short5B = hist5Min[1] > hist5Min[2]

// Plots

// This works
plot_shape_condition = timeframe.period == "2"
plotshape((plot_shape_condition and hist < 0 and go_long2 and go_long5A) ? hist:na, style=shape.labelup, color=color.new(color.green, 90), text="⬆", textcolor=color.new(color.green, 0), size=size.tiny, location=location.bottom)

// This DOESN'T work, doesn't plot any shape at all
plotshape((plot_shape_condition and hist < 0 and go_long2 and go_long5A and go_long5B) ? hist:na, style=shape.labelup, color=color.new(color.green, 90), text="⬆", textcolor=color.new(color.green, 0), size=size.tiny, location=location.bottom)

// These works, so there's data for Hist5Min[1] and [2]
plot(hist5Min, color=color.red)
plot(hist5Min[1], color=color.green)
plot(hist5Min[2], color=color.blue)

这不是plotshape()的bug，是跨周期数据引用的逻辑错误：

• 2分钟周期里，每5根K线对应1根5分钟K线，request.security()返回的hist5Min在连续的2分钟K线上会重复同一个5分钟周期的MACD值
• 当你在2分钟周期里引用hist5Min[1]和hist5Min[2]时，这两个值很多时候和当前的hist5Min是同一个5分钟周期的数据，导致go_long5B = hist5Min[1] < hist5Min[2]的条件几乎无法满足，最终组合条件没有符合的K线，plotshape()自然不显示信号

解决方法是把跨周期的条件计算逻辑放在request.security()内部的5分钟周期环境中完成，而非在主周期（2分钟）里引用历史值。这样能确保跨周期条件是基于5分钟K线的变化来判断，而非重复的2分钟K线数据。

修改后的代码示例：

//////// Long and Short Signals by MACD Histogram, 
//////// 2 minutes(indicator) AND 5 minutes(through request.security()

indicator(title="Moving Average Convergence Divergence Cross w/ Fill & Signal", shorttitle="MACD Cross w/ Fill & Signal")

// 补充原代码遗漏的输入参数
src = input(close, title="Source")
fast_len = input(12, title="Fast Length")
slow_len = input(26, title="Slow Length")
signal_len = input(9, title="Signal Length")
fast_ma_type = input.string("EMA", title="Fast MA Type", options=["SMA", "EMA"])
slow_ma_type = input.string("EMA", title="Slow MA Type", options=["SMA", "EMA"])

// Calculate 2 Min MACD
fast_ma = fast_ma_type == "SMA" ? ta.sma(src, fast_len) : ta.ema(src, fast_len)
slow_ma = fast_ma_type == "SMA" ? ta.sma(src, slow_len) : ta.ema(src, slow_len)
macd = fast_ma - slow_ma
signal = slow_ma_type == "SMA" ? ta.sma(macd, signal_len) : ta.ema(macd, signal_len)
hist = macd - signal

// 在5分钟周期内直接计算多组条件的组合结果
get_5min_long_cond() =>
    [_, _, hist5] = ta.macd(src, fast_len, slow_len, signal_len)
    go_long5A = hist5 > hist5[1]
    go_long5B = hist5[1] < hist5[2]
    go_long5A and go_long5B

// 获取5分钟周期的组合多空条件
5min_long_cond = request.security(syminfo.tickerid, "5", get_5min_long_cond(), gaps = barmerge.gaps_off)

// Go Long conditions
go_long2 = hist > hist[1]

// Plots
plot_shape_condition = timeframe.period == "2"
// 使用预计算好的跨周期组合条件，即可正常显示信号
plotshape((plot_shape_condition and hist < 0 and go_long2 and 5min_long_cond) ? hist:na, style=shape.labelup, color=color.new(color.green, 90), text="⬆", textcolor=color.new(color.green, 0), size=size.tiny, location=location.bottom)

// 验证跨周期数据
[_, _, hist5] = ta.macd(src, fast_len, slow_len, signal_len)
hist5Min = request.security(syminfo.tickerid, "5", hist5, gaps = barmerge.gaps_off)
plot(hist5Min, color=color.red)
plot(hist5Min[1], color=color.green)
plot(hist5Min[2], color=color.blue)

内容的提问来源于stack exchange，提问作者Om Nom