如何为Pine Script直方图的彩色柱状条设置Offset=-2向左偏移

我尝试将代表零线（Zeroline）上下分形的彩色直方图柱状条向左偏移2个单位（offset=-2），该偏移在普通plot中可正常生效，但无法在直方图上实现。我无脚本编写经验，仅会复制粘贴代码及简单调试，已为此卡壳数周，尝试将offset命令应用于直方图相关代码行但无效。

由于Pine Script中plot.style_columns不支持直接使用offset参数，我们需要通过手动偏移数据和分形判断逻辑来实现效果。以下是修改后的完整代码：

//@version=4
study("TSI histogram Fractal Alerts", shorttitle='TSI Alerts', precision=1)

long = input(title="Long Length", type=input.integer, defval=16) 
short = input(title="Short Length", type=input.integer, defval=8) 
sig1 = input(title="Sig1 Length", type=input.integer, defval=6) 
price = close
pc = change(price)
src = input(close, "source")
Zeroline = 0
check = input(true, "Histogram Alert Points")
offset_val = 2  // 定义向左偏移的单位数

double_smooth(src, long, short) =>
    fist_smooth = ema(src, long)
    ema(fist_smooth, short)

double_smoothed_pc = double_smooth(pc, long, short)
double_smoothed_abs_pc = double_smooth(abs(pc), long, short)
tsi_value = 100 * (double_smoothed_pc / double_smoothed_abs_pc)

hist = tsi_value - ema(tsi_value, sig1) 
signal = ema(tsi_value, sig1)

// //////////////////////////////////////////////////////////////
// 调整分形判断的索引，同步偏移逻辑
h1 = hist[4 + offset_val]
h2 = hist[3 + offset_val]
h3 = hist[2 + offset_val]
h4 = hist[1 + offset_val]
h5 = hist[0 + offset_val]

l1 = hist[4 + offset_val]
l2 = hist[3 + offset_val]
l3 = hist[2 + offset_val]
l4 = hist[1 + offset_val]
l5 = hist[0 + offset_val]

histhi = h1 < h2 and h2 < h3 and h3 > h4 and h4 > h5
histlo = l1 > l2 and l2 > l3 and l3 < l4 and l4 < l5

// histogram pattern -----
// 使用偏移后的直方图数据匹配颜色逻辑
hist_offset = hist[offset_val]
hist_up_plus = hist_offset >= 0 and histhi
hist_down_plus = hist_offset <= 0 and histlo
hist_up_minus = hist_offset <= 0 and histhi
hist_down_minus = hist_offset >= 0 and histlo

hist_color = 
   hist_up_plus ? color.rgb(234, 54, 54) :
   hist_down_plus ? color.rgb(39, 154, 89) : 
   hist_up_minus ? color.rgb(67, 85, 187) : 
   hist_down_minus ? color.rgb(183, 172, 50) : color.rgb (242, 244, 244)

// plot -----
// 绘制偏移后的直方图
plot(hist_offset, "Histogram", hist_color, 1, plot.style_columns)
plot(signal, "Signal", #f1cb0a, 1, transp=0, linewidth=1)
plot(tsi_value, "TSI", #0094ff, 1, transp=95, linewidth=1)

// alert -----
plotchar(check and (hist[1] < 0 and hist[0] > 0), "Histogram Crossover", "▔", location.top, color.rgb(247, 225, 61), 20)
plotchar(check and (hist[1] > 0 and hist[0] < 0), "Histogram Crossunder", "▁", location.bottom, color.rgb(247, 225, 61), 20)

plotchar(check and (crossover(signal, +0)), "Signal Line Crossover", "▔", location.top, color.rgb(96, 162, 243), 20)
plotchar(check and (crossunder(signal, -0)), "Signal Line Crossover", "▁", location.bottom, color.rgb(96, 162, 243), 20)

// Line plots
hline(0, title="Zeroline", color=color.rgb(50, 49, 49, transp=50), linewidth=1)
hline(40, title="Overbought", color=color.rgb(36, 140, 36, transp=50), linewidth=1)
hline(-40, title="Oversold", color=color.rgb(197, 49, 49, transp=50), linewidth=1)

hline(20, title="Neutral Zone_T", color=color.rgb(50, 49, 49, transp=50), linewidth=1)
hline(-20, title="Neutral Zone_B", color=color.rgb(50, 49, 49, transp=50), linewidth=1)

关键修改说明 #

1. 统一偏移量控制：新增offset_val = 2变量，后续可直接修改数值调整偏移幅度
2. 分形判断同步偏移：将分形逻辑中用到的hist索引全部加上偏移量，确保分形形态的判断对应到偏移后的柱子位置
3. 直方图数据偏移：创建hist_offset变量存储偏移后的直方图数据，绘制时使用该变量实现柱子左移效果
4. 颜色逻辑匹配：颜色判断使用偏移后的hist_offset，保证分形颜色与偏移后的柱子正确对应

内容的提问来源于stack exchange，提问作者Roadside