Polyline Plus

2023년 10월 30일

This library introduces the `PolylinePlus` type, which is an enhanced version of the built-in PineScript `polyline`. It enables two features that are absent from the built-in type:

1. Developers can now efficiently add or remove points from the polyline. In contrast, the built-in `polyline` type is immutable, requiring developers to create a new instance of the polyline to make changes, which is cumbersome and incurs a significant performance penalty.
2. Each `PolylinePlus` instance can theoretically hold up to ~1M points, surpassing the built-in `polyline` type's limit of 10K points, as long as it does not exceed the memory limit of the PineScript runtime.

Internally, each `PolylinePlus` instance utilizes an array of `line`s and an array of `polyline`s. The `line`s array serves as a buffer to store lines formed by recently added points. When the buffer reaches its capacity, it flushes the contents and converts the lines into polylines. These polylines are expected to undergo fewer updates. This approach is similiar to the concept of "Buffered I/O" in file and network systems. By connecting the underlying lines and polylines, this library achieves an enhanced polyline that is dynamic, efficient, and capable of surpassing the maximum number of points imposed by the built-in polyline.

🔵 API

Step 1: Import this library

Pine Script®

Step 2: Initialize the `PolylinePlus` type.

Pine Script®

There are a few optional params that developers can specify in the constructor to modify the behavior and appearance of the polyline instance.

Pine Script®

Step 3: Push / Pop Points

Pine Script®

🔵 Benchmark

Below is a simple benchmark comparing the performance between `PolylinePlus` and the native `polyline` type for incrementally adding 10K points to a polyline.

Pine Script®

For this benchmark, `PolylinePlus` took ~300ms, whereas the native `polyline` type took ~6000ms.

We can also fine-tune the parameters for `PolylinePlus` to have a larger buffer size for `line`s and a smaller buffer for `polyline`s.

Pine Script®

With the above optimization, it only took `PolylinePlus` ~80ms to process the same 10K points, which is ~75x the performance compared to the native `polyline`.