Trade style specific question [Support and resistence] about ta4j HOT 8 OPEN

Hi.
I just run in same issue, trying find a way how to detect support and resistance level's programmatically.
Did someone has an idea, how other frameworks (might be in python?) will has that or how they do that?

I just will try creating one and give it back to ta4j repo.

Greets,

André.

from ta4j.

Hi.

Cool. good idea! I will do my own research next days too.
Otherwise, i just play around with "Kairi Relative Index" (KRI) which give pretty good "hints" on high's and low's. (Spikes greater 1 for high's and spikes lower -1 for low's)
So, if you interested, i can paste my example code here for research.

KRI Indicator can be found here: https://github.com/Cypher01/ta4j-extended/blob/main/src/main/java/org/ta4j/core/indicators/KRIIndicator.java

Greetings,

Andre.

from ta4j.

from ta4j.

Thanks for the fast reply. So i Look at the Chart and i see a lot of reactions at a specific Level so i draw a box at that Level so that i get the Most reactions but do Not let the Zone get too big.

from ta4j.

from ta4j.

so maybe i make an example to be more clear.

I draw in the chart some support and resistence levels like described here
https://www.thebalancemoney.com/how-to-trade-based-on-support-and-resistance-levels-4043477

and then trade reactions on them based on trends and candlestick patterns.
There are basically some candlestick pattern indicators in ta4j, but i would need to combine them with the support and resistence levels. For these are currently no indicators implemented when i'm right and i am not sure if it is possible to implement something like this into ta4j. What would basically need to be done:

For every bar look into the past and find price points where the chart reacts and create support and resistance points for that. So every bar would basically have something like a set of support and resistance levels because the can change over time. This needs then be combined with the candlestick rules.

from ta4j.

from ta4j.

I'm currently out of the country on vacation until the end of the year but just thinking out loud, perhaps we may be able to implement this as an indicator (i.e. TrendLineResistanceIndicator) that roughly does the following:

1. Finds the last N swing highs (see #1125) using those values as the y-coordinate and their respective index as the x-coordinate
2. Take a "high pass filter" to the results of Step 2, keeping only the highest swing highs (perhaps highest swing high value for every x bars)
3. Uses the values from Step 2 to calculate and cache a modified best fit line (perhaps using SimpleLinearRegressionIndicator or something similar but with a ceiling function)
4. Thus when calling TrendLineResistanceIndicator.calculate(index i) it returns the best fit line values found in step 3 for the given index

Any thoughts on the why or why nots of this approach?

from ta4j.