Hurst exponent lookback too high

the problem

Hi thanks for creating this amazing library! 🙏

I’m trying to use the Hurst exponent on my quotes. I understand that Hurst works optimally when it has 100+ lookback data points. But I want to tweak it for a lookback of 30 lag datapoints since I operate on larger time frames.

https://github.com/DaveSkender/Stock.Indicators/blob/35c46e66a15e7d289a4313f127cc53889bcc6ef4/src/e-k/Hurst/Hurst.Series.cs#L147-L150

• For example in Matlab, it is just a recommendation to have more than 50 data series https://www.mathworks.com/matlabcentral/fileexchange/30076-generalized-hurst-exponent but it not enforced. So once can choose as low as 20
• The Python hurst library enforces the 100 data series, but it has a min_window parameter of 10 https://github.com/Mottl/hurst/blob/5ca5005485a679e6ce11a2769c948915ae27b2da/hurst/__init__.py#L130-L131
• Here is a Simple hurst exponent from Trading View script that allows 30 data points https://www.tradingview.com/script/Sjf9iGCS-Hurst-Exponent-Simple-pig/

Could you please recommend if I should create my own version of if there is a way to disable the validation in this library?

[DaveSkender]

Hi @DaveSkender I have created a PR for your review #978 . Please research and let me know if it needs any changes.

Oh, nice. Thank you for contributing! I’ll take a look and try to review and merge early next week.

[sshquack]

Yes, happy to contribute. Thinking of setting it to 20 since I’ve seen a few python libs have a min of 20 lookback series. I’ll send a PR for you to run a validation after your research.