如何在Pine Script中绘制动态多路径预测股价？解决TradingView线条数量限制难题

Hey there, I totally get your frustration with the line.new limit—270 lines is way over TradingView's 50-object cap, and plot with offset not playing nice with loops is a real pain. Let's break down the best way to solve this using Pine Script's series-based plotting, which avoids the object count limit entirely.

Core Idea: Use Arrays + Series Plotting #

Instead of creating hundreds of individual line objects, we'll store all our predicted high/low values for each path in arrays, then convert those arrays to series that plot() can handle. Each path only needs 2 plot calls (one for high, one for low), so 3 paths = just 6 plots—way under the limit. Plus, we can use fill() to shade the range between each path's high and low for better readability.

Step-by-Step Implementation #

Here's a working example you can adapt to your actual prediction logic:

//@version=5
indicator("Dynamic Multi-Path Price Prediction", overlay=true)

// Configure prediction parameters
prediction_days = 90
num_paths = 3

// Initialize arrays to store predicted values (var = persist across bars)
var float[] path1_high = array.new_float(prediction_days)
var float[] path1_low = array.new_float(prediction_days)
var float[] path2_high = array.new_float(prediction_days)
var float[] path2_low = array.new_float(prediction_days)
var float[] path3_high = array.new_float(prediction_days)
var float[] path3_low = array.new_float(prediction_days)

// Recalculate predictions whenever the spot price updates (last bar only)
if barstate.islast
    current_price = close  // Replace with your spot price source if needed
    
    // Loop through each prediction day to calculate values
    for i from 0 to prediction_days - 1
        // --- Replace this with your actual prediction logic ---
        // Path 1: Bullish scenario (example: +2% daily gain, -1% downside)
        array.set(path1_high, i, current_price * (1 + 0.02) ^ (i + 1))
        array.set(path1_low, i, current_price * (1 - 0.01) ^ (i + 1))
        
        // Path 2: Bearish scenario (example: +1% upside, -2% daily loss)
        array.set(path2_high, i, current_price * (1 + 0.01) ^ (i + 1))
        array.set(path2_low, i, current_price * (1 - 0.02) ^ (i + 1))
        
        // Path 3: Neutral scenario (example: ±0.5% daily range)
        array.set(path3_high, i, current_price * (1 + 0.005) ^ (i + 1))
        array.set(path3_low, i, current_price * (1 - 0.005) ^ (i + 1))

// Convert arrays to series (required for plot())
p1_h = array.to_series(path1_high)
p1_l = array.to_series(path1_low)
p2_h = array.to_series(path2_high)
p2_l = array.to_series(path2_low)
p3_h = array.to_series(path3_high)
p3_l = array.to_series(path3_low)

// Plot each path's high/low and fill the range
plot(p1_h, color=color.new(color.green, 0), title="Path 1 High", offset=-prediction_days + 1)
plot(p1_l, color=color.new(color.green, 0), title="Path 1 Low", offset=-prediction_days + 1)
fill(plot1, plot2, color=color.new(color.green, 90), title="Path 1 Range")

plot(p2_h, color=color.new(color.orange, 0), title="Path 2 High", offset=-prediction_days + 1)
plot(p2_l, color=color.new(color.orange, 0), title="Path 2 Low", offset=-prediction_days + 1)
fill(plot3, plot4, color=color.new(color.orange, 90), title="Path 2 Range")

plot(p3_h, color=color.new(color.red, 0), title="Path 3 High", offset=-prediction_days + 1)
plot(p3_l, color=color.new(color.red, 0), title="Path 3 Low", offset=-prediction_days + 1)
fill(plot5, plot6, color=color.new(color.red, 90), title="Path 3 Range")

Key Details to Note #

1. Persistent Arrays: Using var when initializing arrays ensures they don't reset on every bar—we only update them when the last bar forms (i.e., when the spot price changes).
2. Offset Calculation: The offset=-prediction_days + 1 ensures the first predicted value aligns with the next bar after the current one, stretching subsequent values 90 days into the future.
3. Dynamic Updates: Since we recalculate on barstate.islast, predictions automatically refresh whenever the latest price data comes in—perfect for your dynamic requirement.
4. No Object Limit Issues: Each plot() is a single series-based object, regardless of how many data points it has. Even with 90 days of predictions, you only use 6 plots total.

Alternative: Simplify with Custom Functions #

If you want cleaner code (especially if you add more paths later), wrap the prediction logic in a custom function to avoid repetition:

// Custom function to calculate path values
f_calculate_path(_current_price, _upside_perc, _downside_perc, _days) =>
    var float[] high_vals = array.new_float(_days)
    var float[] low_vals = array.new_float(_days)
    if barstate.islast
        for i from 0 to _days - 1
            array.set(high_vals, i, _current_price * (1 + _upside_perc) ^ (i + 1))
            array.set(low_vals, i, _current_price * (1 - _downside_perc) ^ (i + 1))
    [array.to_series(high_vals), array.to_series(low_vals)]

// Use the function for each path
[p1_h, p1_l] = f_calculate_path(close, 0.02, 0.01, prediction_days)
[p2_h, p2_l] = f_calculate_path(close, 0.01, 0.02, prediction_days)
[p3_h, p3_l] = f_calculate_path(close, 0.005, 0.005, prediction_days)

This makes your code more maintainable if you ever need to adjust the number of paths or prediction logic.

内容的提问来源于stack exchange，提问作者ALEUNG