Greetings -- My newest book, Quantitative Technical Analysis, illustrates the use of both AmiBroker and Python for trading system development. To help show how the two platforms can be used, the same model is implemented for both platforms, with near-identical results. The book continues with additional examples of development and validation using Python. And with coordination between the two platforms. One of the main topics of the book is risk. Beginning with a statement of personal risk tolerance, followed by analysis of the risk inherent in a data series before a model is applied. A major feature of the book is "dynamic position sizing" which determines risk trade-by-trade and recommends position size that will hold drawdown to within the traders tolerance. There is also introduction to trading system development using machine learning, with examples in Python and libraries. The book was published January 23, 2015 and is available on Amazon. Best regards, Howard Bandy
Hi Dr Bandy, Your newest book listed was published 2013. http://www.amazon.com/Dr-Howard-B-Bandy/e/B00MBDSZ3S/ref=dp_byline_cont_book_1 Does your book use Python as a scripting language for Amibroker? Are you making complete systems that are independent of Amibroker? I hope it sells well. A rewrite of your collection to Python would be a great publication. Looking forward to reading it.
Amazon has not updated the page you referred to. Go to Amazon dot com. Do an Amazon search for "Bandy Quantitative" and you will see the newest book. It is blue with a title block set at an angle. It was published two days ago. In it, I show: Complete indicator-based systems in AmiBroker. AmiBroker alone. Complete indicator-based systems in Python. Python alone. Translation of an AmiBroker program to Python, showing identical results. Setting up predictor variables in either AmiBroker or Python for use in machine learning. The process, with examples, of developing machine learning systems in Python. Illustration of 15 different machine learning algorithms (there is not much difference, once the data has been prepared) in Python / scikit-learn. There are examples of communicating between AmiBroker and Python using disk files. That works for any trading system development platform and Python. There are no calls to Python from AmiBroker. Or vice-versa. Additionally and equally importantly, the book shows: How to define personal risk tolerance. How to evaluate the risk inherent in a data series before any model is applied. How to evaluate the risk in a trading system. How to set position size trade-by-trade so the drawdown does not exceed personal risk tolerance. It is called "safe-f". How to define and use a universal objective function. It is called "CAR25". It is a risk-normalized forecast of account growth. The book has its own website. www dot QuantitativeTechnicalAnalysis dot com Look at the /book page. There are several chapters that can be downloaded and read. Best, Howard
As I take it and by reading their forum AmiBroker seem to release a Python add-on in one of upcoming weeks/months.
I am not an employee of or consultant to AmiBroker or privy to the details of their plans. That said, the chatter on the Yahoo AmiBroker Forum, including some comments from Tomasz Janeczko, president of AmiBroker, suggest that there will be an AmiBroker -- Python interface that is supported by the AmiBroker organization.
I am surprised your not, Dr. Bandy. Your book is the most sought after by Amibroker developers. I don't even know an alternative. Perhaps, more collaboration with T.J. is in order.
Nope. Nothing beyond a shared beer when we are both at the same conference. I pay retail for my AmiBroker license.
For whatever it's worth, it's my belief that adding a python add-on shows a lot of prescience. I think it's a smart move as it will make their product more accessible to a broader audience of interested traders. Also, Python is similar to R and MATLAB in that it can be used somewhat interactively for research purposes as well, so maybe those interested in this approach will find it easier to bridge the gap between researching an idea and creating a tradable system. I'm not an AmiBroker user and likely will not be, but I'm generally in favor of almost any change that will contribute to the (pardon my cliche) "democratization of automated trading." I think it's a smart business move on their part.
All the book's content, all to the dot, can be easily extracted for free off the net. I am happy to provide links to various topics, ask and you shall be given, but I strongly advise not to feed snake oil sales men (who do not even pay here to spread their advertising) who basically copy cat information from public domain and attempt to sell books to newbies who just do not know better. * Amibroker website offers tons of free materials, certainly instructions how to set up indicator-based systems * There are numerous Python libraries that offer testing and implementation of trading strategies. A simple web search takes you straight to it * No translation between Python and Amibroker needed, for what??? And if you need it you need to re-write the code in the same way than any other language transfer works, the two languages are NOT compatible and I am willing to bet my life that the book will not provide any shortcuts * Machine learning is the most ridiculous term introduced into retail finance and algo trading since its inception. I am willing to argue my point but in summary Amibroker as well as a simple loop in Python lets you iterate parameter values and in that way (curve)fit. Genetic algorithms are offered in Amibroker as well as in numerous Python libraries, again, nothing special here. Scikit is an open Python library and there are tons of Youtube videos that walk anyone through the steps. * diskfiles to "communicate" between Python and Amibroker??? Dude do you even know what a garbage you are saying? * All risk tolerace terms and concepts, mentioned, are blatantly stolen from other books or authors. I hope you have given credit and asked for permission otherwise you may have some lawsuits up your arse. In any case no novel material, all simple basic stuff that is openly available on the net. * No respect for your PhD (or Dr. whatever) for this garbage, pal. Seriously, the equivalent is to sell rotten tomatoes to blind people.