Hey guys I want to write my own stock screener + data storage & analysis + alarm in python and run it on a server and want it to notify when alarms are triggered either by text message or E-mail
I have one concern right now:
So far I found two solutions for python that make the connection to IB API easy, IBridgePy (ibridgepy.com
Now my suspiciousness kicks in. Why are they offering their software for free? There is no ad, no monthly cost, no nothing
Is there any way they might have built in a backdoor that sends them information from all the scripts made and then they use the sucessful ones for themselves or even selling them and eventually rendering them useless?
Is there any way to find out? (Wireshark?)
IB API for stock screener
Other urls found in this thread:
ibridgepy.com
code.google.com
interactivebrokers.com
twitter.com
anyone used/using IBridgePy personally?
shameless self bump, if you guys help me out I'll start tripping & share with you the next moon missions if backtesting is successful
I like that picture
Because it's just a wrapper over a C++ library. You can check out more here; ibridgepy.com
Plus I'm assuming they make money from their "rent a coder" section of the page where they offer their services for algo trading.
The actual python wrapper they built isn't all that crazy imho.
There's even an official version: code.google.com
yeah I read that but somehow I need 100% trust, know any good program that would monitor the software for a period of time and report any suspicious connections?
ibpy =/= IBridgePy
also if you download the zip folder there's a .py file called "quantopian". any idea if they have anything to do with quantopian? Because Quantopian, for example, is openly using and cashin on people's algos
also IBridgePy claims it's completely open source but the wrapper is a .pyd file that is not decompilable so how would I know what's all in that file?
or am I getting something wrong?
any help appreeeciated
will 100% share moon missions till the SEC gets me
$ file IBCpp.pyd
IBCpp.pyd: PE32 executable (DLL) (GUI) Intel 80386, for MS Windows
It's a windows library, so it won't even run on linux.
I'd try first the official api since it has python support interactivebrokers.com
I have no idea how the IB api works, but is there no way to create an user/api key with read only access?
I'm running windows, there is no linux version for IBridgePy
I downloaded their API and they claim they have sample code for python but I can't see it in the folder I downloaded, only everything else they said they have
Yes it's possible to have read only access for the api
but my concern is that the dll might send my code/strategy somewhere where it could be evaluated and used in case it's successful lol, I'm wearing a tinfoil hat right now but im serious, I want some proof that this wouldnt happen
I just saw that Python is only available in the newest beta version of their api. Just downloaded it. Thanks for the input! trying to figure out if it works the same way as IBridgePy
this is awesome, just got released a week ago
New Python API
Note: Requires Python 3.1 or greater.
Beginning with release 9.73, a new Python API client is now included. After you install this beta release on your computer, you can find Python API components in the following locations:
Python API sample code – located in the samples/Python folder in your API installation directory (typically IB_973)
Python source code – located in the source/pythonclient folder in your API installation directory
im getting class import errors all the time. didnt know python was case sensitive
fixed that but there seem to be privilege problems all the time, gotta phone IB to ask them what version of python they used for their beta and what OS
this is the kinda scholarly discussion i come here to see man
lets go and build an API in python, man. where do I begin?? how do i do this
Sounds cool OP I have been fucking around with quantopian for the last month, but I also want to make something that's stand alone.
So I'm also looking for API's and other ways to get data. First I'm following some machine/deep learning tutorials to see if it could be usefull in my algoritm/script.
Atm I'm following a youtube tutorial by a guy named sentdex.
If you want to keep in contact OP, maybe start a discord, let me know.
just learned that IB has their own python API in their newest beta release now
For data collection IBridgePy offers a google and yahoo market data collector which you could use if you want free data
for real time data you need to open an account with IB, i think it's the best broker out there to trade, if you just want market data id go with tdameritrade/thinkorswim but i wouldnt trade for 10 bucks a trade
IB requires $3000 to open acc if you're under 25 but you can withdraw it again as far as I know
I learning myself right now, mainly teaching myself python
nice, how was quantopian?
thanks for the yt recommendation, but personally im teaching myself the simple stuff first, python
I'll make a thread again with same pic in 1-2 once i progressed with learning python
Machine learning is actually pretty simple, basically just some applied statistics. Deep learning and neural nets are a bit more complicated, but have some easy tools if you only want to use it. Still have to dive deeper into those.
All the tutorials in the yt reference I gave you are in python btw, so you can learn in parallel if you want haha.
> how was quantopian
It is a little weird getting used to it, it's still weird to write everything in methods.
But it is an enourmously powerfull tool, especially for starters. The API (data from morningstar) is totally free and has insane fundamental data, 8000 companies and over 600 metrics per company.
But you can't use it outside of quantopian I think.
The graphs and analytics tools are also nice.
>But you can't use it outside of quantopian
that's a big deal breaker
but fortunately IB finally has their own python API now (aswell Java, C#, C++, DDE etc)
Did you manage to write any successful trading algo in quantopian?
>succesfull
I guess, I was trading above the market at a pretty stable rate, long term.
But yea, no fees, no pressure, it's just a simulation.
I know I wouldn't be comfortable using that algo if I was trading irl.
It was just an exercise to get used to financial programming in python and the basic tools.
So I'm almost ready to make something real (stand alone) and add some innovative stuff besides the basics.
learning python 101 by Michael Driscoll now, it's a very good book so far