Question: support and resistance levels ?

[firewater]

Wondering if ta-lib-python has the ability to calculate support and resistance levels for a stock given open, high, low, close, volume data. Thanks.

[mrjbq7]

What is the calculation in particular that you are looking for?

[firewater]

Thanks for the quick reply.

I'm used to determining support and resistance by eye and was hoping to find a Python library to help automate the process. I don't quite understand the mathematics involved, so I've provided a couple resources below that could help develop a functional calculation.

PivotPoints @ Investopedia.com
This link recommends using pivot points to find support and resistance levels and provides a formula:

Perhaps that's a good starting point ?

These next two links point to Reddit where a user was also wanting to programmatically calculate support and resistance levels.
Suggested options include the Donchian Indicator, Volume Profile Point of Control (superior to pivot points?), and more. Honestly the number of options are overwhelming:

• https://www.reddit.com/r/algotrading/comments/18bevqj/support_resistance/
• https://www.reddit.com/r/algotrading/comments/nl35ye/how_to_calculate_support_and_resistance_levels_in/

I wish I could provide more definitive information to help build support and resistance into ta-lib-python. Hopefully the information above provides a foundation on which we can build. Please let me know if there is anything else I can do.

[mrjbq7]

And perhaps this is useful for that:

https://wire.insiderfinance.io/pivot-point-calculation-using-python-1bd1e75f17e6

[mario4tier]

A known alternative worth mentioning: Donchian Channels — the rolling N-bar high, low and
midpoint. It is the usual way to get support/resistance levels without any calendar awareness: it
takes plain high/low arrays, no timestamps and no notion of "previous day/week". This one is
planned for TA-Lib — tracked here:
TA-Lib/ta-lib-proposal-drafts#38

Pivot points are still being looked at (drafts#20), but are less certain.
The classic form is anchored to the previous calendar period, which TA-Lib's function API has no
way to express since it receives only OHLCV arrays and no timestamps.