Data Feed with API for Order Book Analysis

[natedredd10 10]

I do wonder if you really are familiar with these tools (to be honest it doesn't sound like it)? Your "waiting 10-20 years" comment is nonsense, R is open source so it's all available now and guess what? ... if you use a licensed copy of Matlab MathWorks will provide details of algorithms and even chunks of source code.

 

As I asked before I am interested in what you are using Matlab / R for?

 

I'm not waiting 10-20 years per se to trade....from what i know right now it will take 10-20 years to use these AMAZING tools properly to AUGMENT my "intelligence" and add algo EV to my discretionary EV ..A decade to show someone like Ditmar "whats up"...fuck "AI"..even though the concept of augmented intelligence would be the same acronym...

I will admit that if there is utility at all in a moving average model, i believe there is more utility in a hidden markov model, more EV, just because of how absurd simple models are compared to modern markets...As if the entire field of time series analysis died 40 years ago.. Of course there is the pit of "i have a phd in x, not many people understand x, x is somewhat correlated to markets, leverage x model...load the gravy train.."

Fuck them, I want to exploit them too.

My argument for matlab R is really much simpler though..Rentech is hardly in the business of burning cash..Rentech employs Leonard E. Baum...

Baum

If you trade ES for a period of time, bet your ass a part of villain is Leonard E. Baum trading ideas, massively capitalized trading ideas....

basically, nothing is stopping me from upgrading my stick in a machine gun fight beyond my own lazyness and lack of hard work...

As if market profile is a good weapon against an information theory god, backed by the best hedge fund ever...

As if we are so dumbass we can't think of something better than 20 yo market profile bullshit..its just lazy.

[cunparis 10]

I have a few ideas for things I'd like to do and I am debating between trying to write a ninjatrader indicator or using the zenfire API directly. Blowfish - how is it going with the zen-fire API?

 

I think they have a java version, i plan to try that. I'm a java guy, good at OO but not very familiar with C#. I got a bit discouraged reading the posts about the zenfire API and think maybe the limitations of a ninjatrader indicator would make it easier & faster. At least for prototyping ideas.

[become 10]

cunparis,

 

I developed some things with the Zenfire API for a while when I was especially interested in the trade intensity stuff. In the end I decided it wasn't worth all the extra work (and I didn't get into charting etc). I found I could do everything needed via NT (gomi's work helped too).

 

(fwiw)

[BlowFish 10]

I have a few ideas for things I'd like to do and I am debating between trying to write a ninjatrader indicator or using the zenfire API directly. Blowfish - how is it going with the zen-fire API?

 

 

I have to regretfully say currently it it is not. I got to the point where I was getting data into a grid and doing some basic manipulation.

 

Whilst I can see the advantage of a managed framework (C#) it really seems to require a disproportionate amount of energy to learn it's in's and outs.Having said that writing in other 'environments' learning the 'environment; is usually the greatest challenge (whether this is directly to the Windows API, the Ninja API or a Java VM for example).

 

If it was me I'd be inclined to do it in the easiest 'environment' (even if just for prototyping) unless that imposes limitations you find unacceptable.

[cunparis 10]

I have to regretfully say currently it it is not. I got to the point where I was getting data into a grid and doing some basic manipulation.

 

Whilst I can see the advantage of a managed framework (C#) it really seems to require a disproportionate amount of energy to learn it's in's and outs.Having said that writing in other 'environments' learning the 'environment; is usually the greatest challenge (whether this is directly to the Windows API, the Ninja API or a Java VM for example).

 

If it was me I'd be inclined to do it in the easiest 'environment' (even if just for prototyping) unless that imposes limitations you find unacceptable.

 

Thanks for the feedback. I'm used to programming for Ninja and it's quite easy (not as easy as EasyLanguage but still their API is logical and the documentation is good). Yesterday I was playing around with the OnMarketData & OnMarketDepth methods in Ninja.

 

One of the limitations I was hoping to overcome is the fact that Ninja uses 1 second timestamps while I believe zenfire uses millisecond.

 

I don't have a lot of time for programming. Mainly when the markets are slow at lunch, dinner, and weekends. So I agree on taking the easy route.

 

I am impressed that some companies are writing to the zenfire feed directly. OFA comes to mind. And there are no charges for the API. So far it seems developer friendly, but as you said, complex & time-consuming.

[FulcrumTrader 10]

The only real information that can at times help set you up for potential directional changes in price is Order Book Delta (from all the un-traded order activity in the order book). I do have an Order Book Delta tool that I run on my NT7 to identify ACV ratios of 2:1 or greater. If all you are going to do is run ACV ratio analysis in the real time order book activities you should be fine with just pulling data from Zenfire API or TT Fix feeds that are available. If you are going to do more intense analysis then you probably will need to go TT API or CQG.

 

Many traders from my group will all be out in Vegas the 17th to 20th of November this year. We are having a Cumulative Delta based traders gathering to look at some of these exact Transaction Level Analysis issues collectively (these are the same dates the Las Vegas Online Trade Expo is going on at Caesars Palace). Those interested in Cumulative Delta and related work will be working together in one of my houses to see what tools we can create for better analysis or automation. We will also be looking into several other specific areas of interest;

 

CD Volume to Total Volume Ratios

 

CD Momentum

 

Trade Intensity

 

CDVWAP Indicators

 

Order Book Delta Indicators

 

We have a good diverse group that will be looking into all these subjects to see what good ideas we can generate as a focused group (programmers and traders will be out in Vegas for this gathering). And of course as a group in the evenings we will be out counting cards as a combined Blackjack hit team to make millions........JK.......LOL! Any of you Cumulative Delta junkies want to meet up one night in Vegas during these dates in November just let me know......we have several scheduled fun times set up in the evenings we can all meet somewhere!

[cunparis 10]

Any of you Cumulative Delta junkies want to meet up one night in Vegas during these dates in November just let me know......we have several scheduled fun times set up in the evenings we can all meet somewhere!

 

If I were in the US I'd love to join you but it's quite far from Europe. Sounds like a lot of fun though.

[FulcrumTrader 10]

We will most likely run another Cumulative Delta based trader gathering in Paris in very early August.....I have many EU contacts that want an EU event so those are the dates we are looking at. Also, a Miami Beach in the spring event may take place first....we are looking at dates at this time. Miami will be the "Beach, Babes, and Bots" event!

 

JK....LOL!

[cls71 10]

Hello,

 

maybe my experience could be useful for someone ...

 

I'm working on analyzing the Level2. I had problems with the Level2 of NinjaTrader + Zenfire (however the Level1 was fine). I could see that not all events were processed. So I decided to work with Rithmic.

 

I have developed a tool from scratch to see the market delta (Level1 analysis) and I am now developing another tool for analyzing the Level2.

Everything is programmed from scratch, using Rithmic's API and VisualStudio (.NET and C#).

 

I have not had any problems with the multithreading.

 

Rithmic events come with microseconds

 

Regards