Hello Nicolas , I am new to proreal time , its a great platform and it seems the support base is very good asswell its good to see everyone helping each other to achieve there goals .I am going to start trying to build my own indicator and learning the code , before I start I’d thought of asking if even possible . I am trying to get a MACD over a stochastics chart , basically a MACD with a 80 and 20 line . Any input would be appreciated . Thanks Matt
So you want to normalize the MACD, because this indicator don’t have a fixed scale since it is calculated from moving average differences.
This is of course possible but to normalize it on a percentage scale, like the stochastic, we must define first a lookback period to determine the top of bottom of the MACD values over this period and then dynamically calculate the current percentage of the MACD from it. Are you OK with this idea?
Hello Nicolas , Thanks for your response , yes I am OK with this idea , although I have no idea of putting it all together . I have started trying to learn how to modify indicators and learn the code , it might take a while . I suppose the bigger the look back period the more accurate or true the indicator would be ? What would you suggest as a look back period . Thanks again Matt
I was trying to say that the MACD information will not be the same once this modification will be made because an MACD don’t have a fixed scale and that’s what your are trying to do here. A stochastic has a scale of 0 to 100% while a MACD is a difference in points.
Hi Nicholas, hope you are well. Please can you provide a sample of how the macd normalisation code would look like? Thanks.
To normalize a data serie, we must define first a scale. As an example, this MACD indicator is normalized with its Z-score : MACD Z-Score standardized value
Just wondering why the OP requires this and what he is trying to define here ? Any indicator i ever write has a tangible definable measure/filter of a price action/pattern . If we have an idea of what the filter is trying to achieve other alternative may be suggested . I am just failing to see where this objective is an advantage .. Maybe im blind but as a long term technical trader a squilley line isnt much use to me if i dont know whats being represented . It’s all great writing code but i like to know what its use is before i start . Normalization is a useful tool but in some cases it detracts from the original objective of a filter . Extremes of some indicators is erased by normalization and in many cases extremes are where the best opportunities lie . I am a price action trader and a coder second and as such are less likely to data mine , trust me i am still guilty of mining . The point here is what is MACD , MA easy sure but convergence and divergence are almost immediately removed by normalizing ?? Correct me if i am wrong ??
Ultimately instead of turning MACD into a normalized indi just using one of many already normalized indis may be a viable altermative . Knowing the objective an alternative can be easier to suggest
