Welcome to NexusFi: the best trading community on the planet, with over 150,000 members Sign Up Now for Free
Genuine reviews from real traders, not fake reviews from stealth vendors
Quality education from leading professional traders
We are a friendly, helpful, and positive community
We do not tolerate rude behavior, trolling, or vendors advertising in posts
We are here to help, just let us know what you need
You'll need to register in order to view the content of the threads and start contributing to our community. It's free for basic access, or support us by becoming an Elite Member -- see if you qualify for a discount below.
-- Big Mike, Site Administrator
(If you already have an account, login at the top of the page)
Broker: Mirus (Broker), Continuum (Data), Dorman (Clearing)
Trading: Futures
Posts: 202 since Mar 2013
Thanks Given: 428
Thanks Received: 202
--
I'd like to thank you for pulling together an incredible swing indicator. My hat is off to you.
I'd also like to thank everyone else for contributing thus far.
Over the past month or so, I built a swing indicator that is based off of Larry Williams market structure that he presents in his book "Long Term Secrets to Short Term Trading", 2nd edition, 2012. It's pretty similar to what you have done, but with a few minor differences. Williams defines short term ("ST") points in the following manner:
The Standard Swing Type gets pretty close to these definitions. However there is one small difference that is key. For a ST swing point to be valid, Williams states that in the case of a ST high that the low of the ST high bar must be breached lower before becoming a valid ST high. Similarly on the short side, the high of the ST low bar must be breached higher before becoming a valid ST low. I was able to do this via lists, but I'll be damned if you are not a fine programmer who has a much deeper grasp of the .NET language's capabilities. So perhaps you have some ideas here.
So although the Standard Swing Type using a 1 bar Swing Size gets us close to how Williams define his ST swing points, it's not perfect. Using a 2 bar Swing Size also gets us close, but not quite. But I think in this case, I've decided not to let the good be the enemy of great. If you ever wanted to collaborate further on incorporating Williams ST swings into the code, I'd be glad to help out (just PM me).
However the ST swing points really end up being noise when they are set to 1 bar. But here is why that doesn't matter. Williams goes on to define intermediate term ("IT") and long term ("LT") swing points in the same exact manner that he defines ST swing points.
And as you guessed it, you can then zoom out even further to get an feeling of where you sit on the long term scale through the following definitions of LT swings.
So here is the true value of Williams' view of market structure: It is ENTIRELY objective. There is no subjective variable that the trader can manipulate with their inborn biases. It is 100% reliant upon what the price action has produced. For this reason, I love it.
I also like his interpretation of market structure because it's transferable between different markets and time frames, and in one chart you can see where you stand on the ST, IT and LT swings. So for instance, you might only take longs off the ST or IT swing points if the last LT swing to print was a LT low. And instead of having a 1 week long term chart open assessing the long term swings (and hoping that you have the right Swing Size in place), a 1 day chart open assign the intermediate term swings (and again, hoping you have the Swing Size right) and a 60 min chart that you use for entries and exits and to assess the short term swings (and again, crossing your fingers that you are using the correct Swing Size), you just have one chart with all three swing points on it so that in a glance you can see where you stand on all swing horizons.
I'd love to collaborate with you on incorporating Williams ST, IT and LT market structure as one of the Swing Type options. I love how it objectively determines trend without allowing any room for us to mess with settings, while showing ST, IT and LT market structure on one chart.
Additionally, based on my study of Williams market structure, I also saw the value in collecting swing statistics to provide some statistical context for whether I should or should not take a given swing trade and how many bars and ticks have been observed by this particular type of swing in the past. I came across post here (
Hi Guys. I think this indicator deserves it's award and would like to express my sincere thanks to Dorschden for this indicator. I'd also like to share some research which maybe of interest to some people to show other ways of using this indicator.
I've …
), which is just freaking incredible. If you haven't read this post, read it. He gets it.
It looks like DJ exported the various swing information to the Output Window (via Print commands) and then copied into Excel, at which point he did his analysis. In my indicator, I had been looking at the same concepts (prints, Output Window, copy, paste into Excel, analyze), while also looking at MAE per swing to get a bit of context on risk. But then I saw on your latest edition this Statistics option and the resulting tabs that pop up on the bottom of the screen, and I immediately concluded that I could just create a new tab showing DJ's stats and some other ones that I have been looking at lately--and then everyone could see what DJ and I have been seeing. These stats really provide excellent context, and they have helped me in the development of a tradable edge.
So I'd like to build this new tab, and I'll PM you the code when I'm done for your review. And then perhaps you could incorporate the new stats tab into a new version for everyone. If there are any specific statistical observations that anyone wants, please let me know and I'll scratch my head and see if I can't get them in there.
To complete the stats tab, I needed to confirm what variables in the Base indicator you've used to collect the various swing relation counts. I've taken a stab at what I thought the variables where, but I couldn't get them to add up to your upCount and downCount variables--so I know that I'm missing something. Therefore, could you please help me understand your count variable naming convention? I'm assuming that "hhCountLL" means the number of swings that have gone from a HH to a LL. Is this correct? Or does ""hhCountLL" mean the number of swings that have gone from a LL to a HH? Once I understand this piece, I should be able to get the counts to square away and look like DJ's Excel work (here https://nexusfi.com/attachments/62806d1329345829).
Also, from my cursory review of your ticks by swing type, it doesn't appear that you have broken the ticks and bar durations down by swing type yet. Is that correct? If that is right, I'm assuming that I'll need to create some new variables to hold these values, which I will create per your naming conventions. Please let me know if I have missed anything. Perhaps these values are already being stored, in which case all I need to do is collect them and format them in the new tab.
Again, this is an excellent piece of work. Thank you and everyone else (@Silvester17 has made some excellent contributions to this thread) for all of the work you have done thus far. I am hopeful that I can help by assisting with the incorporation of the Williams Market Structure concepts, as well as with expanding the stats information.
I'm trying to plot Swing High and Low Values on my chart to resemble FT71's Harmonic Rotations using my Ninja Charts. I currently have the Price Action Swing Indicator Pro on my chart,but it appears that it is not capturing all swing High and Low points.I'm not sure if it is a setting parameter that I have wrong or this may not be the best choice for Harmonic Rotations.
Broker: Mirus (Broker), Continuum (Data), Dorman (Clearing)
Trading: Futures
Posts: 202 since Mar 2013
Thanks Given: 428
Thanks Received: 202
Howdy--
So I spent a bit of time augmenting the PAS code to add a third tab (indexed to position 4) and populating my first stab at the columns and rows.
I have numbered the rows to aid in indexing, however the numbering will eventually go away.
I'm assuming that hhCount is equal to the count of all HH, that hlCount is equal to the count of all HL, hhCountHL is equal to the count of all HH to HL swings, lhCountLL is equal to the count of all LH to LL swings, etc. Can someone please confirm theses assumptions are correct?
Again, I'm shooting for the tab to look like @djkiki's work:
I have numbered my percentiles, counts and bars to the 0.05, however I might truncate this down to 0.10 if the 0.05 resolution is too fine (i.e., not enough variation).
I have found that it does not appear that there are counters from swing points to the DT and DB, so it looks like I'll have to create these counters.
Does anyone know off the top of their head which variables I would access to populate the mean function. I'll basically need to sum all of the ticks and divide by the count. So for instance, sum all of the HH to HL ticks and then divide by the count of the HH to LL swings. Does anyone know which variables I would access to do this?
I've snooped around, and it looks like I'll need to drop all of the swing tick movements into a sorted array so that I can find the midpoint (i.e., median). So I'll just copy what others have done in the C# world on that.
For the Standard Deviations, I plan on doing a SD(1) and SD(2), so that we can get some sense on the variability of the data. I fully expect SD(2) to be larger than the mean, in which case I might drop it. However I'll give it a shot. I'll see how this goes. I'm hoping that the help I can get on the mean process will inform the SDs.
And same on the Percentiles, Counts and Bars...hopefully the mean process will inform me there.
I previously posted a question above reguarding FT 71's Harmonic rotation swings,but am having greater problems setting up my PAS on my 8 Range NQ chart. I am able to get some Swing High/Low Labels,but that seems to be it.I can't seem to get any Double Top/Bottom,HH,HL,LH,LL,Labels...I tried changing the parameters but no success.I'm also trying to incorporate the Divergence study using GOM CD,not sure what I'm doing wrong.I have been going over all the posts in this thread trying to figure out these issues but with no avail.I would greatly appreciate any help.
Hi sorry I have no clues but do realize that along with your request/dilemma some images of what you speak of help those that are willing to help much more.
(there are some amazing people here that can figure out just about anything) Thanks
e.g. images of your chart, indicator settings panel etc.
along with some annotations for each. Along with, even if previously posted, an explanation(including images) what it is you wish to accomplish
the pas indicators all are working great with range charts, including the gom cd divrgence:
made an excellent point about showing your chart, settings etc. so for now I can only guess. one guess would be you're using a dark blackground and the text color is black.
Thanks for your quick response.I am attaching a few screenshots of my NQ 8 Range chart and the Parameters that I currently have set,I'm obviously doing something wrong with the parameters because my chart doesn't resemble yours.I am also trying to get my GOM CD to work with the PAS to show divergences.Thanks Again for all your help,I truly appreciate your time.
first you don't have the "gom cd" features enabled in your pas pro indicator. I attached a version where gom cd is enabled. copy that version in your indicator folder and click "copy and replace" when a window pops up.
also it looks like you don't have any data recorded. your gom cd indicator shows a straight line. if you don't have any data recorded, the pas pro is not able to show any cd divergences.
I also attached a chart template with the settings I used in my previous post. you can use that to get started and then change it the way you like it.
Check out comments starting with post #757 thru 762 starting on page 76 of this thread. There are some initialization and round off issues with feature counting.
Just saw your post...Awesome !!! Thanks so much for the Chart templates and parameters,I'll be working on it later today...Thanks again...Have a Great weekend !!!