Anyone up for collaberating on replicating Chan scripts in Python?

Quote from tickzoom:

To make it easy to quickly "implement" a trading idea with python (I know python) that will serve a group of people (not just yourself) requires enormous work.

It's because you need basic plumbing to load and process the data, synchronize it for multi symbol portfolio trading, handle generating bars in multiple time frames, handle processing of various types of orders (simulated unless you're using tick data). If using tick data , you have many months of work to get the performance down to anything acceptable.... the list goes on and on and on.

I want to propose to you a possible simpler and quicker option to use python.

have you heard of IronPython? It's a free python implementation which runs on .NET instead of the python parser. That means it's fully compatible with C#, VB, etc.

That also means it can interact with already existing C# objects and so on.

Well, some people have suggested adding python to the tickzoom project.

The best way would be to do it using IronPython.

tickzoom already has all those facilities and plumbing above (and many more) for historical testing and easy switch to live trading.

No actual paying user of tickzoom as requested python but it's on my list only because I find the idea very interesting personally.

If you're at all interested in doing this, I have a free trial for tickzoom to let you look into it.

If you think you can do it, I'll give you the exclusive opportunity to using the free trial as long as you need to get IronPython working. If you get it working, I'll give you a free license so you can run it live.

I think the market for this financially is very, very small--almost non-existent. I'm only offering this because it's very interesting technically and "geeky" so I'll be happy to support you however I can to do this.

If you want to discuss further, just sign up on the waiting list (on my signature) and reply to the emails--that's my real personal email used for the list. I don't get PMs or email notification on elitetrader.

Wayne

Thanks for the opportunity; I'll keep it in mind. Although, I hadn't planned to carry it that far (full blown real time)... just yet. I'm still very much a neophyte to python.

My main objective was for this to be an experiment where participants had an opportunity to learn how to do something (simple backtesting etc...) like this with an open source language.

Thanks again, for the feedback. I'll try to get a look at iron python when I get a chance.
 
Crude and small-- long only script. Execution time to fully load csv and return generated results and plot noted on bottom. Seems a little slower than meas results(I assume time is in microsec default units; I'll have to go check).

29gmq1s.jpg


Going to start getting more sophisticated soon. Took about 2-3 days to code. Way faster to prototype than some other languages.:D
 
Glad to see you made some quick progress.

Do you have an initial impression yet on how much quicker/slower it is to develop ideas in Matlab vs. Python?

Eric
 
Quote from Trader922:

Glad to see you made some quick progress.

Do you have an initial impression yet on how much quicker/slower it is to develop ideas in Matlab vs. Python?

Eric

Hey Trader922,

Honestly, I haven't used matlab professionally for years. So from my current perspective, python seems more flexible (lower language).:D
Don't recall doing much OOP in matlab, but python has both OOP and matrix based features. Also, the industry didn't have that much package development back then, so matlab was a must for stuff like DSP.

Matlab does seem to have a lot more readily built functions available, however. Like kalman filter and cadf (from 1st ex), for instance.
 
That's funny, because unless I made a mistake (which I'm doubting) there are some huge mistakes in the code and very basic metrics as I continue along. Makes me wonder if he is truly using these scripts to back-test. And they are not just typos, they are fundamental concepts. Anyone actually take the time to run all the excel/matlab scripts? Catch any? PM me if you did. I'd be interested to compare.
 
Quote from dtrader98:

That's funny, because unless I made a mistake (which I'm doubting) there are some huge mistakes in the code and very basic metrics as I continue along. Makes me wonder if he is truly using these scripts to back-test. And they are not just typos, they are fundamental concepts. Anyone actually take the time to run all the excel/matlab scripts? Catch any? PM me if you did. I'd be interested to compare.
I'm not surprised at all. Many in the know have slammed his book - but lets just leave it at that...
 
dtrader, I just ran across this book the other day that might be of interest to you..
Financial Modelling in Python by Shayne Fletcher..
I would think it might contain some of the grunt work that you are looking for although from the table of contents its focus is on derivative pricing models.
As someone who is trying to learn matlab though, I do have to ask what do you gain from python that would rather not just use matlab straight away?
 
Hi Jdeezero,

Already ordered! Although, thanks for the heads up. I could have used matlab, although, I prefer to use python for a few reasons. One, is I have some custom blocks I took time to build in python that are not available in matlab. Two, I mentioned I like using a lot of these open source languages as the package development is tremendous (even beginning to exceed commercial products in some areas, like statistics). Lastly, it feels more like a programming language than matlab (i.e. more flexible); at least that's been my experience so far.

Lower level languages like java and c variants, also seem to lack some of the package development I mentioned, as well as take longer development time.
Python also works well with other languages, as a prior poster mentioned. So in many ways, python is ideal (and free to boot).
 
dtrader,

I believe I recall you used rapidminer at one point as well. Are you still using it? If not, are you using something other package for data mining or using python instead?


Quote from dtrader98:

Hi Jdeezero,

Already ordered! Although, thanks for the heads up. I could have used matlab, although, I prefer to use python for a few reasons. One, is I have some custom blocks I took time to build in python that are not available in matlab. Two, I mentioned I like using a lot of these open source languages as the package development is tremendous (even beginning to exceed commercial products in some areas, like statistics). Lastly, it feels more like a programming language than matlab (i.e. more flexible); at least that's been my experience so far.

Lower level languages like java and c variants, also seem to lack some of the package development I mentioned, as well as take longer development time.
Python also works well with other languages, as a prior poster mentioned. So in many ways, python is ideal (and free to boot).
 
Quote from bozwood:

dtrader,

I believe I recall you used rapidminer at one point as well. Are you still using it? If not, are you using something other package for data mining or using python instead?

Hi. I use several (including custom), although I'd prefer not to divulge specifics. Rapidminer was capable and a program I shared the time to pick up (on the other thread) and offered others to participate, but ultimately, like this thread-- lots of eyeballs, not much participation.

Conclusion on RM: too much of a GUI hog/lack of flexiblity IMO.
 
Back
Top