For me, ‘instead’ of coffee is not an option …

… an interesting article ‘and’ a cup of coffee can make a stimulating Sunday morning. See here for medical details:

http://penguintraders.com/forums/topic/the-take-a-break-from-serious-business-thread/page/3/#post-9645

Best,

s.

And one more thing to mention: in the source code posted by G. a link to a conceptual article by Waleed Aly Khalil was included:

http://www.ifta.org/public/files/journal/d_ifta_journal_09.pdf

I already downloaded that IFTA Journal more than a year ago, but only regarded a John Ehlers article worth reading in detail. Not a good choice, right?

s.

Less than 2k of code … best things in life seem to follow the KISS rule!

Thanks @tankhang & @gg53.

s.

Great job – thanks for that:

I assume your analysis is based on classical OHLC of M1 and not on tick data, and the logic of up / down volume is the same as discussed here – correct?

In my opinion the results described in your last paragraph are the most interesting ones. In your definition of range5 = close-open do you mean the signed value or…[Read more]

Sorry guys, didn’t make it yesterday! Found a bug late in the evening. This indicator requires ticks coming in to work, so debugging will be done on Monday.

I’ll post my code after debugging.

s.

@Anti:

Hope you’ll stay tuned! You don’t have to be an experienced coder to contribute valuable conceptional thoughts. And maybe your R skills and my mql skills could be an interesting combination. I’m still thinking about that possible volume vs. price movement correlation. Analyzed in R and coded in mql this could be pretty powerful.

And…[Read more]

In my intrabar volume indicator, a new unit ‘points per tick’ is introduced, see bottom indicator in pic attached.

Also extending the time window is possible now by input parameter. The pic shows an example of 5 minutes history at 1 sample per second.

Sample rate can be reduced now, default is still 1 bar per second.

Indicator will be published…[Read more]

G:

Bars disappear because I’m currently watching a rolling time window of 64 seconds, sampled 1 bar per second. When a certain bar has passed this window from right to left, it will disappear.

As I stated: this window could easily be extended if a larger sample makes sense. Would cost some additional cpu and memory resources and lead to longer…[Read more]

G, following this statement of yours …

… would you say it makes sense to take my prototype posted here and use it as a blueprint to exploit tick data further?

At the moment, this prototype indi is nothing more than a proof of concept that recent tick data can be stored in a data structure alongside with price data and very basic stats…[Read more]

Thanks for clarifying:

So, in this model, zero efficiency (doji) is represented by a value of 0.5, while max. efficiency (no wicks) by a value of 1.0.

When dividing body size by bar range, zero efficiency is 0.0, while max. efficiency is 1.0.

Both scales appear to be linear, so there’s not much of a difference.

s.

Vlan:

Look here for my newest blueprint – might save you some time to manage that session averaging.

It’s still a prototype! If you find bugs I’d appreciate if you notified me.

Have fun!

s.

Ok,

Following G’s repeated proposal (thanks for that!) of averaging volume per session, I now attach a first simple prototype of that kind.

For now no gimmicks, no normalization (except for time), no oscillator. Unit shown is ticks per second, bars colour coded in a simple manner for up / down.

By default, session averages are being calculated…[Read more]

Thanks for your feedback!

Yes:

I do understand your assumption, but I just don’t believe that it’s true and I also don’t believe that ‘this was shown before’, as you’re stating.

So for your formulas: uppath / totalpath would be kind of a ‘bar efficiency’, equal bodySize / barRange. Correct?

s.

@Anti:

Could you provide some conceptual background about this special part of the algorithm:

if (Close>Open)
buff=((High-Low)*MathPow(2*High-2*Low+Open-Close,-1))*iVolume(NULL,0,i);
else if (Close<Open)
buff=((High-Open+Close-Low)*MathPow(2*High-2*Low-Open+Close,-1))*iVolume(NULL,0,i);
else
buff=iVolume(NULL,0,i)/2;


s.

Great:

But could you anyway try on your old laptop and report here? Would save me the trouble to grab some age-old laptop, install MT on it and check for myself!

s.

Yes:

That’s probably the case. I only wanted to provide an example of how the indicator looks like when no ticks are coming in.

s.

Ok, and now for the indicator.

I still consider this one experimental, a prototype and possible basis for further intrabar development.

@Anti: you wrote that you’re working on some weak laptop at the moment. So your feedback about ressource consumption would be highly appreciated. On my XEON system CPU consumption lines hardly move when I launch…[Read more]

Yep, I watched in my new magnifying lens I coded last night (see pics):

Pic 1: normal activity. I chose AUDJPY for this example for minor influence by EUR and USD news G. mentioned.
Pic 2: during the event I had a phase of more than 1 minute with no ticks coming in (broker GP)
Pic 3: ticks are coming in again
Pic 4: up to 800 points per…[Read more]

Yep, I watched in my new magnifying lens I coded last night (see pics):

Pic 1: normal activity. I chose AUDJPY for this example for minor influence by EUR and USD news G. mentioned.
Pic 2: during the event I had a phase of more than 1 minute with no ticks coming in (broker GP)
Pic 3: ticks are coming in again
Pic 4: up to 800 points per…[Read more]