BlowFish

Not sure you really need a loop at all. Take another look at http://www.traderslaboratory.com/forums/56/building-volumeprofile-indicator-easylanguage-7073-2.html#post79526 If you go for fixed scaling all you need to do is re-plot the appropriate histo bar of the array element you have just updated.

I would disagree. Do you think someone bidding for several millions (or 10's of millions) of an instrument over days, weeks or even months, is not 'above' some one buying a 100 shares? What is 'the real auction'? and how is the method of buying and selling relevant to the analysis of a time series?

Yes, maybe one for shoulda woulda coulda Having said that this sort of action is fairly common and so perhaps interesting as a general approach to 'backing and filling' action.

No contracts in spot just a minimum amount of currency to trade http://individuals.interactivebrokers.com/en/trading/exchanges.php?exch=ibfxpro

I would say NO to be honest. Far too easy to get seduced and distracted. It's difficult enough to purposely move towards your goal without adding distractions. There is another big issue and that is 'grailittis' if one has difficulty trading its much easier to swap indicators/systems than to change your own behaviour. With pure PA there is much less to blame. That's not to say all indicators are useless but they present a greater risk than there potential reward. Of course if you have trouble telling up from down (:D like Kiwi:D:D) then by all means slap on a moving average.

It is with extreme prejudice that I refuse to read anything with Elliot in the title however, as its recommended by you I'll make an exception Actually the trouble with Elliot is he got his basic building blocks wrong, if you break things down further and then build everything from ABCD moves you can make a completely consistent (and so potentially useful)l framework.

It might be worth pointing that a lot of 'backing and filling' can effect the R:R. For example if you are looking to get long after a lazy move down. Here is a reasonable case from the end of last week, the EURGBP. I show a couple of possible entries and a couple of possible stops. You can see because of the nature of the swings stops are rather further than targets. I'd be interested how people might have managed this? It is worth mentioning that this occurred at a good potential support level (imho) so a place where you might want to get long.

There is a VPOC indicator in the indicators section. It only plots the peak volume however (not a full histogram). If you look for PVP or Jperl market statistics you should have no trouble finding it.

One thing that seems plausible to me is that if markets are in fact multi fractals (do not have a single fractal dimension) that trending markets and stationary markets (cycling or range bound if you like) possibly have different fractal dimensions. I wonder if separating the cyclical components and the non-stationary components (a la Ehlers) and then modelling those separately might bear fruitful results. To be honest this was just a gut feel from someone (me) that is not at all well schooled in the mathematics. Intuitively it makes some sense.

Any order that you absolutely want executed should end up a market order when elected. Whether using stop losses, emergency stops, trailing stops or whatever as part of your money management scheme is a good thing is a slightly different debate. If you want to absolutely be sure your orders are filled when elected then you should not use a limit.

I didn't answer when you first posted mainly because I didn't have much to offer that could not be found with google. There have been a couple of discussions here though they often tend to descend into the 'are markets fractal' debate.

Range charts should be identical provided you start plotting them from exactly the same place each time. (If you think about it the first tick will be your reference point). Some implementations might reset at midnight or session start time giving some small discrepancies. As with all charts 'tuning' them will depend on what you want to see. What you want to see should depend on what your objectives are (e.g. the size of swings you are looking to capture). There is no one size fits all.

You could contact them and see if they run any programs to train people on there products but ultimately you will need access to the infrastructure that the API's run on. Again maybe they offer packages for independent developers?

Quite, I have a handful of his books and actually mentored with him many many years ago. The law of charts is free and a pretty good resource, I too would be hesitant to recommend the higher price works without reservation. For example one area it (TLoC) may help people is in early (and consistant) identification of side ways action. This is what I figured which is why I asked the question rather tentatively. JR does present a couple of ideas that I would be interested to hear your opinion on of course should I ask about those I will ask in the context of what you have written here. I'll hold fire for now anyway I just started reviewing it from the beginning (though I have followed it from the start) I suspected something was different or had evolved. I was wrong. This has turned out to be a far sleepier thread than I thought it would be. But reading charts from the "hard right edge" as they say is what this business is all about, so I'll carry on for a bit longer. -: made me smile. I hope you are pleased that you carried on! The ego needs a degree of complexity to get 'fed'. If things are really that simple it tricks us into thinking there must be more or can even blind us to what is obvious as a last ditch resort! Me too - current favourite.

TT I wondered whether you where familiar with Joe Ross' Law of Charts? It is essentially 'price action' with all the usual 'patterns' HH HL HH...HH HL LH etc. Certainly don't want to lead anyone off track so I'll say no more at the moment.

All good point's.I used to drive myself crazy trying to 'finesse' stops with theses considerations, but there is no 'perfect' solution. (with trading there never is).

If it is only updating a couple of times a day thats all fine and dandy! but for building accurate volume profiles you need to be updating the arrays every tick. Horses for courses I guess For example because TS and MC dont have 'tick precise' technology where you can build indicators tick by tick on historical data I will simply use (for example) a week of single tick's to build a weekly picture. (I do this for VWAP and SD's sometimes) and also PVP. An alternate way would be to use say 2 minute bars and only update every 2 minutes. You would average the volume from high to low of the 2 minute bar. This will give quite different results for the peak volume than a tick by tick approach which is why I developed this code. Of course whatever works for you! But I can say all the existing VWAP and PVP code essentially locks up on a lot of bars! Maybe it would recover eventually when it catches up but it is horribly inefficient. Even the vol histogram code in the first post can take minutes to load on a couple of days of minute bars.

I have the Jeremy du Plessis one, it is as good as any and being a more modern book does at least touch on intraday P&F too.

I'll certainly be interested to see how things take shape. May the momentum always be on the side of your trade:)

Loops are very important and powerful constructs however one should be aware they can be quite slow as all the code in the loop is repeated for each iteration. This is particularly so if you have nested loops (loops within loops) or many iterations particularly on indicators that update every tick. Always keep the code in loops to the bare minimum. Novices often put all sorts of things that could go outside the loop on the inside Often there is a way to do things without using them at all. e.g. using the original example. HighPriceSum = HighPriceSum + High - High[9]; Will return the same value though it should be noted that this is 'unstable' for the first 9 bars on the chart (mind you so is the other looping code). Both Tradestation and Multicharts do a pretty decent job of detecting how many bars are needed before an indicator 'works' and so there is no need to worry about this (unless you are a purist).

As with most things there is an upside and a down side to trading rooms. If someone is having difficulty focusing joining a room is unlikely to help. (I guess it may do if peer pressure can be brought to bare). If someone is focusing but 'missing' stuff then their approach is perhaps too complicated or perhaps there method is not rigorously enough defined. Where rooms shine (well for me at least) is helping with the boredom and solitary aspects of trading. I also find it interesting what other people are looking at (though potentially distracting). I wonder why you retired from trading? perhaps something to do with needing 'eyes in the back of your head' or general fatigue? I guess you are looking to return or you wouldn't be posting this thread? I dunno but for me better solutions to the eyes in the back of the head issue or fatigue etc.) has been simplify (and/or clarify) my approach. Look at less instruments or 'zoom out' of the action (look to trade larger/longer swings). Actually for some (the more hyperactive) zooming in might be effective in keeping you focused. guys making 100's of trades per day tend not to be so easily distracted! I can't help feeling if one is missing stuff they are trying to look at too much. Having said all that a room is a good place to remind you that there is major news coming up, or we are at 5 year highs, or it's triple witching today etc. etc. Or even general observations about S/R or other well watched technical stuff (oh, we are sitting on the 200 day moving average). Things you might not notice or even actively watch but that is of general interest, or even an excuse for a long lunch!

If I could manage to get whet you manage to out of excel I'd probably do it like that too!

Actually my code is not intensive at all. Arrays are absolutely fine looping thorugh them isn't. I would go out on a limb and say there is no more efficient way to do things in El to that which I presented. None of the code I have submitted has a single loop in it except to initialise all data to zero. The sugestion i have made for histogram has not a single loop either. I also have non iterative code for VWAP & weighted SD's (many orders of magnitude faster than tradestations own, they get exponentially faster the more bars you load) I decided not to share that as I believe it has commercial value (not that I have exploited that). Take a look I think it is a very elegant aproach (though of course I am biased).

Actually my code is not intensive at all. Arrays are absolutely fine looping thorugh them isn't. I would go out on a limb and say there is no more efficient way to do things in El to that which I presented. None of the code I have submitted has a single loop in it except to initialise all data to zero. The sugestion i have made for histogram has not a single loop either. I also have non iterative code for VWAP & weighted SD's I decided not to share that as I believe it has commercial value (not that I have exploited that). Take a look I think it is a very elegant aproach (though of course I am biased).

Well there is a way to avoid rescaling. Lets take the case that does need rescaling first. If you look at GKMarketProfileTL (this is from dim distant memory) there is a parameter for number of bars that the profile occupies. In other words there is a fixed space on the X axis of say or 100 'bars' (whatever you enter for that parameter). So the profile occupies first bar of the chart to the 100th bar of the chart. This means that whenever the peak profile gets longer you need to rescale with respect to that maximum value. This is acceptable as it only occurs if the peak surpasses its old value. An alternate way is that you have a parameter, N bars (or pixels or charcters) per X volume. That is an absolute scale. So for the ES you may have 1 'bar' (or ascii char) per 250,000 contracts. Using this method your chart simply 'grows' from the left. Every time the peak volume increases by 250,000 contracts you expand your chart 1 more 'chunk'. Of course this has the potential problem of the profile growing bigger than the screen but you could have a simple test to see if that has occurred and truncate it. The user would see that the bars are occupying the whole screen and could then manually change the scaling to say 1 'bar' per 400,000 contracts (or whatever). I like this approach.