ninja trader good or bad

Yes, OQ is a lot better but not set up for the non programmer to learn easily like Ninja. I can't get answers to things via their 3 different documents and the forum sometimes.... today I'm realizing i have to answer some of my own questions by experimentation, I have to stop everything when I have a question and write up some code to output things to the console to tell me what the program is doing, and that in realtime sometimes so questions have to wait for real time hours... it's slow going for me but I do have a strategy that runs for backtests, it needs just a little more polish and some fixes so that it works in realtime the same as backtesting [the bracket orders get rejected by the TWS simulated account for some reason and I don't know why yet] and I am up and running after only about 6-7 weeks with the software... With easy language I probably would never have gotten to this point, the event driven framework is just so way better than anything else. I was never comfortable with Ninja, I did experience some problems with manual trading from the chart that were puzzling and the way the thing checks in to the license server and more at startup made me uncomfortable about protection of my intellectual property, with OpenQuant I can run it behind a whitelisted firewall that only allows the computer to connect to interactivebrokers.

Sky, maybe you know, I use the OnOrderFilled event to place the bracket OCA order with code copied from Dr. Anton's post regarding placement of the bracket, the prices are in line with the markets and exactly as intended but for some reason IB's simulated account rejects them.. I just discovered this this morning. I found examples of the ProviderError code on the foruns just now, probably I can experiment and find out what TWS is doing but maybe this is a problem you've seen?
 
Hi,

there is a Provider Errors window in OpenQuant. You should be able to see all error and other messages coming from IB TWS in this window. You can also click on an order in the OrderManager window and see reject reason in the Execution Report window.

Regards,
Anton
 
Quote from maxpi:


Sky, maybe you know, I use the OnOrderFilled event to place the bracket OCA order with code copied from Dr. Anton's post regarding placement of the bracket, the prices are in line with the markets and exactly as intended but for some reason IB's simulated account rejects them.. I just discovered this this morning. I found examples of the ProviderError code on the foruns just now, probably I can experiment and find out what TWS is doing but maybe this is a problem you've seen?
More...

TWS requires that the orders have 2 decimal places use Math.Round(Price, 2); ~ that could be your error, it was mine
 
Thanks much.....

I knew about that and I managed to get it right for the entry order but not the bracket order!!!!!! :(

I worked with a guy that was a programmer, his career went back to vacuum tube computers, he was a very rare breed indeed. He said that you need a huge ego to program because the computer will tell you how incompetent you are many times per day!! I believe him!! :eek:
 
Quote from maxpi:

Yes, OQ is a lot better but not set up for the non programmer to learn easily like Ninja. I can't get answers to things via their 3 different documents and the forum sometimes.... today I'm realizing i have to answer some of my own questions by experimentation, I have to stop everything when I have a question and write up some code to output things to the console to tell me what the program is doing, and that in realtime sometimes so questions have to wait for real time hours... it's slow going for me but I do have a strategy that runs for backtests, it needs just a little more polish and some fixes so that it works in realtime the same as backtesting [the bracket orders get rejected by the TWS simulated account for some reason and I don't know why yet] and I am up and running after only about 6-7 weeks with the software... With easy language I probably would never have gotten to this point, the event driven framework is just so way better than anything else. I was never comfortable with Ninja, I did experience some problems with manual trading from the chart that were puzzling and the way the thing checks in to the license server and more at startup made me uncomfortable about protection of my intellectual property, with OpenQuant I can run it behind a whitelisted firewall that only allows the computer to connect to interactivebrokers.

Sky, maybe you know, I use the OnOrderFilled event to place the bracket OCA order with code copied from Dr. Anton's post regarding placement of the bracket, the prices are in line with the markets and exactly as intended but for some reason IB's simulated account rejects them.. I just discovered this this morning. I found examples of the ProviderError code on the foruns just now, probably I can experiment and find out what TWS is doing but maybe this is a problem you've seen?
More...

The lack of documentation on the quant platform is very concerning. I looked and don't see much other than a blog of tutorials and a quick start guide.

I would like to hear from anyone that has experience with both quant and ninja, or another c# based platform. Looking for specific examples.
 
actually you really can find what you need in the forums. I have experience with both and I have lifetime memeberships for both. NT is a good platform, with good support. OQ is a much better platform with okay support
 
Quote from Sky123987:

actually you really can find what you need in the forums. I have experience with both and I have lifetime memeberships for both. NT is a good platform, with good support. OQ is a much better platform with okay support
More...

The support really is ok. So far I have been able to do what I want to do but it feels like a game of rumpelstiltskin a lot of the time....... I simply hate C language, I had a class in it years ago, it was the most horrendous struggle to learn something I ever underwent and I'm phobic about the language, i sort of dread working with it every time I start.

I absolutely would not go back to any software that did not have the event driven structure, and I don't know of another one that does. I was reading on the forum some thread where somebody suggested that another event was needed and the team put it in right away... kudos to the team for that kind of responsiveness and ability to be flexible and listen to the users and writing code that can be added to and maintained and all, they are smart people there with the ability to design something right at the outset, I've done a zillion engineering projects and that ability to get things right at the outset is uncommon..... I worked for a consulting guy once that managed projects, he was making thousands per day because he could take a room full of corporate leaders and get them to tell him exaclty what they wanted us to produce, that could take as long as two weeks sometimes, and then he would set up a contract for the consulting work that made it very very costly for them to try to make changes during the project. We produced all the state of the art disc drive designs for years [IBM, Toshiba, Goldstar, and alot of other ones, we designed many of their drives] because we could get it all right at the outset and nobody could screw our life up during the project... you can't do that in the corporate environment, the sales department will be demanding changes all during your project and the engineering department won't be clear on the goal at the outset.. in a way working in Openquant is like working for that consultant. Your projects start out with a damn good framework and changes are not a problem, you find out which event you want to run the new code in and just install the code...
 
Anton,

I have heard many people speak very well of OpenQuant, but nearly all cite the lack of documentation as a serious concern.

I'm a fair programmer and I would really like to take OQ for a spin. I believe I might not be hindered by some of the limitations I experience with NinjaTrader. However, the reality is that without the necessary documentation, it's too big a commitment of time for me to try to begin using it.

Do you have specific, near-term plans to add such documentation? This might attract a new group of paying customers to OQ.


Regards,
 
Quote from JangoFolly:

Anton,

I have heard many people speak very well of OpenQuant, but nearly all cite the lack of documentation as a serious concern.

I'm a fair programmer and I would really like to take OQ for a spin. I believe I might not be hindered by some of the limitations I experience with NinjaTrader. However, the reality is that without the necessary documentation, it's too big a commitment of time for me to try to begin using it.

Do you have specific, near-term plans to add such documentation? This might attract a new group of paying customers to OQ.


Regards,
More...

Probably you should go ahead and start up with OQ, if you're a programmer you won't have that much trouble, the documentation for a not-so-great-hacker-actually programmer is sketchy compared to Ninja Trader, it's not a show stopper.... they document their whole API in detail, it's just that they stop there, if they followed up with code examples for everything in the API document, right in the API document like Ninja has, anybody could learn the software in no time..... I was surprised that I could pick up on Ninja Trader programming so rapidly, it was all about the code examples, I could start every project knowing that I could finish it in good time, not true with OQ, I can start knowing that I can finish but the schedule...........
 
I've been playing with NT recently evaluating it against OQ which I'll play with more later.

So far NTs good points are that:

- it's relatively easy to set up and get started, seems well designed in general
- it's relatively quick
- good charts, lots of built in indicators
- free to use until you want to live trade
- supports the brokers/feeds I want
- c# based so it should be possible to do almost anything with some wrangling

So far the bad is:

- As far as I can tell it doesn't automatically support a non-intraday fully automated system. You have to write something to keep track of where you are in a strategy.
- It's easy to backtest against a basket of stocks but not easy to run a strategy against a basket. You have to 'run' each strategy on each symbol individually which is a hassle if you want to run a strat against 100 symbols (for example).
- You can't create a strat using the wizard that mixes indicators of different bar types (like using SMA(10day) in a strat that is running on 1min bars) everything runs on 'bar' time which is an option that is global to the strategy (apparently you can get around this if you write NinjaScript directly though)
- The work cycle is tedious, create a strat in the wizard, save it, open strategy tester, pick the instrument, run the strat, decide to change something, to back to step 1 and repeat...I like to test ideas quickly and this work cycle is slower than what I would like. I want to be able to make a change and press F5 to run my new strat against some set of instruments.
- Others have mentioned that you can't do advanced things like route your order based on conditions. I'm not at the level where that matters but I can see it being a liability in the future.

dislaimer: These are just my first impressions based on a limited amount of experience. There might be ways around all of the cons listed above that I just don't know about yet.
 
You must log in or register to reply here.
Back
Top