Parabolic SAR

[real-lazy-coder]

Parabolic SAR (PSAR) Type: Trend

The Parabolic Stop and Reverse (PSAR) is a technical indicator that identifies potential reversals in the price movement of an asset. It appears as a series of dots placed either above or below the price, depending on the direction the price is moving. The indicator is used to determine the direction of a trend, identify potential entry and exit points, and set trailing stop-loss orders.

Checklist for new indicators

• Indicator is documented and implemented by extending BigIndicatorSeries
• A "faster" version of this indicator is implemented by extending NumberIndicatorSeries
• Tests for getResult are present
• Tests for highest and lowest result caching are present (if not already existent for base class)
• Tests for error handling are present (if not already existent for base class)
• Indicators (standard & faster version) are exposed in src/index.ts
• Indicators (standard & faster version) are added to startBenchmark.ts
• Indicator is listed in README.md

[bennycode]

@real-lazy-coder brilliant PR! 👌 Thanks for doing all the work. It is just missing some test coverage. Have you tried EarlyAI to generate test code? Or the Cursor AI Editor?

[real-lazy-coder]

@bennycode I'm on the struggle bus. For the life of me I cannot figure out how to get 100% test coverage on the PSAR.

 ParabolicSAR.ts         |     100 |    97.67 |     100 |     100 | 108,152   

These are edge cases and I'm at a loss. Any guidance?

[real-lazy-coder]

@real-lazy-coder brilliant PR! 👌 Thanks for doing all the work. It is just missing some test coverage. Have you tried EarlyAI to generate test code? Or the Cursor AI Editor?

I'm using Claude Code quite a bit these days. But it's 🔥 my 💰 at a blazing fast rate!

I'll look into EarlyAI. Thanks,

[bennycode]

These are edge cases and I'm at a loss. Any guidance?

I ran the tests and reviewed the HTML output in the /coverage directory. It looks like we need to cover scenarios where it falls back to previousLow (line 108):

sar = previousLow.lt(sar) ? previousLow : sar;

Similarly, we should add a test case where it falls back to previousHigh (line 158):

sar = previousHigh.gt(sar) ? previousHigh : sar;

[bennycode]

@real-lazy-coder I was able to write tests for the missing cases -> real-lazy-coder#1

Example:

it('adjusts the SAR when the previous low is lower than the pre-previous low', () => {
  const psar = new PSAR({accelerationMax: 0.2, accelerationStep: 0.02});
  const prePreviousLow = 8;
  const previousLow = 7;

  psar.add({high: 10, low: 9});
  psar.add({high: 11, low: prePreviousLow});
  psar.add({high: 12, low: previousLow});

  expect(psar.getResultOrThrow().toString()).toBe('12');
});

[bennycode]

@real-lazy-coder shall we get this in? 😀

[real-lazy-coder]

I have 1 last line I need to write a test for. I'm currently on vacation with my family. I will get this finalized when I return next week.

[real-lazy-coder]

I have 1 last line I need to write a test for. I'm currently on vacation with my family. I will get this finalized when I return next week.

[bennycode]

I have 1 last line I need to write a test for. I'm currently on vacation with my family. I will get this finalized when I return next week.

Sure! Take it easy and enjoy the vacation with your loves ones. 😊

[bennycode]

Hey @real-lazy-coder, can I support you to get the PR done? Hope you had a nice vacation! 😀