Software products from Piligrimm

 

Software products from Piligrimm

_______________________________________

In this topic, I want to start the presentation of the programs I develop for trading. To begin with, I will present five indicators, in the future I will exhibit other developments of indicators, expert and trading systems. Before proceeding to the description of indicators, I would like to briefly explain the principles underlying my developments.

I started modeling and forecasting time series in 1978. As a basis for his work, he used the Method of Group Consideration of Arguments, this is something close to modern genetic algorithms. For 10 years of work in this direction, he has worked out his approaches to the analysis and forecasting of time series in conditions of strong noise, and a large percentage of the random component in the original signal, as well as the synthesis of additional signals based on the input, in order to increase the information content of the input signal and improve the quality of the forecast. In 1988, he left work in this direction and went into business, doing it until 2000. Since 2000, he has been researching the Forex market and developing programs for its analysis and trading. I mainly use Matlab as a basic package, I started studying MQL 4 a year ago, however, I still haven’t used it much, only in writing indicators that collect initial data for calculation and display charts on the terminal screen, I carry out all calculations in Matlab. After many years of studying the Forex market and a large number of experiments, I came to the firm conviction that it is quite possible to create effective predictive expert systems for its analysis and making trading decisions. It is possible to do this most effectively on the basis of multivariate data analysis using neural networks. As input data for neural networks, in my opinion, it is most effective to use data synthesized or filtered from a multicurrency signal that has undergone preliminary processing in order to increase information content. The expert systems themselves should be built on the principle of self-learning and self-organizing systems, continuously adapting to the changing market situation. In this direction, I am leading my developments. Additional information about my approach to market research and program development can be found in the topic: “Effective trading strategy based on multi-currency analysis of several DCs” https://www.mql5.com/ru/forum/103138 , as well as from my article: “Is it possible to predict the Forex market? How to create your own trading strategy? http :// articles . mql 4. com / ru /372 . True, I signed this article with the pseudonym DAO . This is due to the fact that a year and a half ago I first came to this forum and, having seen the topics discussed here, I decided to write an article and share my experience, hoping that it might be useful to someone, that a discussion would start, and others would express their vision and understanding of the Forex market. But there was no discussion, and soon I stopped following this topic. Under the name Piligrimm , I asked questions on the forum regarding problems related to mastering MQL 4, but gradually got involved and began to participate in the discussion of other topics.

____________________________________________


Indicator on Wavelet-transformations "Kristi_GrafWav".

I want to start the presentation of indicators with the "Kristi_GrafWav" indicator.

Indicator "Kristi_GrafWav" designed for multi-currency analysis for three instruments. Currency groups, time frames are selected at your discretion. Charts will be displayed in one window at the same scale, you can choose any of the 13 charts that are created both for individual instruments and for their combined use. The indicator is intended for use both for manual trading and in conjunction with Expert Advisors to create Trading Systems, but was developed as a tool more focused on manual trading, as well as market research in order to select the most effective group of tools that complement each other when designing multicurrency trading systems. systems.

The first version of the indicator used exe - a file written and compiled in Matlab.

I already put it up for sale, but sales experience has shown that many of the users do not want to download an unknown exe file to their computer - a file with a large number of auxiliary libraries necessary for its operation. I redesigned the indicator, transferring the entire calculation to the indicator written in MQL 4, and the wavelet transformations are supplied as a source code written in Matlab, and the user himself uploads it to Matlab and compiles it on his computer, creating the exe file necessary for the indicator to work. The package includes detailed instructions, with the help of which, even a user with no experience with Matlab, can easily do everything. Work example indicator can be seen in the figure below, the instruments EURUSD, EURGBP, EURJPY are used. Trading decisions based on this indicator can be made at the points of change of direction or intersection of some signal lines.

Cost of the "Kristi_GrafWav" indicator: $85 US .

______________________________________________

_______________________________________________


Indicator "Kristi_GrafModelWav"

With the "Kristi_GrafModelWav" indicator, I begin the presentation of a group of 4 indicators that complement each other and are designed both for manual trading and for creating trading systems. The indicators are implemented in MQL4 as standard custom indicators.

Indicator "Kristi_GrafModelWav" is designed to form a group of 40 signals synthesized on the basis of polynomials, and increase the information content of the instrument used for trading. Polynomials are trained to simulate the work of wavelet transforms, the learning algorithm was set in such a way as to create a spectrum of signals with different phase - frequency characteristics, but which are derivatives of the instrument on which the indicator is installed and carry its features. Such a variety of signals allows, when fed to the input of the neural network, to obtain a sufficiently informative input sample for modeling various processes, as well as predicting the direction of the trend. And although some of the signals are visually similar to each other, the polynomials on the basis of which they are obtained differ in their characteristics and were trained using different algorithms, and these internal differences are perceived by neural networks.

When developing a trading system based on this indicator together with a neural network, I decided to change the parameters of one of the two very similar to each other, to make them more different, as a result, the accuracy of training the neural network dropped by an order of magnitude. I returned everything to its original form, considering that higher accuracy is better than visual perception of diversity.

The indicator has the ability to normalize the output signals in a wide range by setting the normalization coefficient in the settings.

If necessary, by setting the appropriate resolution in the settings, the file with synthesized signals can be saved to disk for further processing by other programs. Polynomials were trained on quotes of currency pairs, but the indicator also works on other instruments, not only the Forex market, although I did not conduct a detailed study of its work on other markets. An example of the indicator operation is shown in the figure.

Cost of the "Kristi_GrafModelWav" indicator: $150 US .

_______________________________________________

____________________________________________

_______________________________________________


Indicator " Indicator Trend"

This indicator was developed as a signal source for neural networks, against which the network is trained. Its readings can be viewed as a simulation of one of the harmonics of the trend. The principle of operation of the indicator is similar to that of ZigZag , but unlike it, in addition to local minima and maxima, intermediate trend reversal points are created corresponding to a change in the intensity of its movement. This solution significantly increases the information content of the "Indicator Trend" indicator signal compared to ZigZag , and this significantly affects the accuracy of neural network forecasts when using this indicator. The use of this indicator is effective not only in conjunction with neural networks, but also directly in conjunction with an Expert Advisor, as well as for manual trading. Saving the signal values calculated by the indicator to disk, with appropriate permissions in the settings, is possible in two ways:

one). Saving signal values on each bar;

2). Saving signal values only at trend reversal points.

The signal values at the last two breakpoints are also stored in global variables, the names of global variables can be changed, this allows you to set any number of indicators with different threshold levels on one instrument and organize communication with an Expert Advisor or other indicators through global variables. When the indicator is working, it is possible to optimize its parameters by rebuilding the threshold coefficient. The choice of a specific threshold value is determined by on which instrument and timeframe the indicator is installed, as well as what trading strategy you want to implement based on this indicator. The indicator works in two modes, static and dynamic. In static mode, the indicator builds a trend model for the entire length of a given history sample and recalculates it for the entire length upon the arrival of a new bar, creating templates of a given length that can be presented as a training sample for a neural network. In dynamic mode, the indicator calculates the value only on a newly arrived bar, gradually accumulating the calculated values as new data arrives. An example of the indicator operation is shown in the figure below, indicator "Indicator Trend" corresponds to the red line of the chart.

Price indicator "Indicator Trend": $150 US.

_______________________________________________

"Indicator Power" indicator

The idea on the basis of which the indicator " Indicator " was developed power " is to dynamically monitor the struggle between the bullish and bearish trends that are taking place in the market at the moment. When the bullish trend dominates, as new bars arrive, the indicator values on the zero bar rise in proportion to the strength of the trend, and when the bearish trend dominates, they go down accordingly. With slight fluctuations in the trend, then in one direction, then in the other direction, the indicator values are reflected in the form of a straight line, which along its entire length from the last turning point slowly shifts in proportion to the strength of the dominant trend, and with a significant increase in any of the trends, it abruptly moves in the corresponding direction , forming a new breaking point. The readings of this indicator give an objective idea of what is currently dominating the market, a trend or a flat. If a as new bars arrive, the indicator draws a straight line or steps that are small in level and rare in time, this indicates a flat condition, but if transitions from one level to another become frequent and dynamic, this indicates a transition to a trend. What is considered a trend and what is a flat, everyone determines for himself in accordance with his ideas about the market, and can be adjusted in the indicator by changing the threshold level in the settings. As in the previous indicator, in the " Indicator Power " it is possible to save the calculated values to disk for all bars on the length of the LengthSample sample, which is set in the settings, and the values of the last two breakpoints - in global variables. At high threshold levels, the indicator " Indicator power " can be used as an indicator of a market phase change. When developing trading systems, many are faced with the problem that the trading system at different phases of the market often shows opposite results, making a profit or completely draining the deposit. In this regard, the question of determining the moment of change of these phases becomes very relevant. Practical experience of working with this indicator has shown that with the correct selection of the sample length and threshold level, it reflects the change in the market phase in the form of a sign change between the difference between the two values of global variables given by this indicator. By analogy with the "Indicator Trend" indicator, the " Indicator power " works in two modes, static and dynamic. An example of the indicator operation is shown in the figure below, indicator "Indicator Power" corresponds to the pink line of the graph.

The cost of the indicator " Indicator Power ": $150 US .

____________________________________________ _____

"Indicator Channel" indicator

" Indicator Channel " creates a trend channel and reflects the dynamics of its behavior, extrapolating the direction of the market trend development. The informative components of the indicator are: narrowing and widening of the channel, angles of divergence of the upper and lower beams, inflection points and moments of reversal of the central line of the channel. When new bars arrive, the indicator dynamically reflects the changing market trend, rebuilding the rays from the last breaking point to the zero bar. As in the previous indicator, in the " Indicator Channel " it is possible to save the calculated values to the disk for all bars within the length of the LengthSample sample, and the values of the last two breakpoints for the upper, middle and lower lines are stored in global variables. An example of the indicator operation is shown in the figure below, indicator "Indicator Channel " corresponds to the blue and yellow lines of the graph.

Price indicator "Indicator Channel": $150 US.

_______________________________________________



_______________________________________________


An example of building a trading system

To a greater extent, the proposed indicators were developed for use in trading systems. As an example of using four indicators: "Kristi_GrafModelWav", "Indicator Trend", "Indicator Power", " Indicator Channel ", I wrote a trading system, hoping to demonstrate its work with indicators in real mode on a demo account. The trading system used two different strategies using a different combination of indicators to manage two orders. This was done in order to cover the possible losses of one order with another. The trading system was created to work with EURUSD H 1. Unfortunately, the trading system turned out to be very slow, making an average of 1.5 trading operations per week. I had to greatly coarsen the system parameters in order to eliminate false positives. The results of testing the trading system for half a year are in the attached archive. Joint testing on two orders - Strategy Tester, separate testing on the first order - Strategy Tester1, separate testing on the second order - strategy tester2. No optimization of the trading system was not carried out. With such a slow work, a visual demonstration of this trading system would not work.

I decided to make another trading system for EURUSD M 1. Having set up the trading system initially for Sell for one order, I got about the same trading indicators as for the first trading system. About 50% of net profit per month with a drawdown of about 8% for one Sell trading signal when trading with a constant lot. But the peculiarity of this trading system was that in the favorable phase of the market, it made a series of trading operations, when an unfavorable phase of the market occurred, it entered into a state of expectation and could not trade for 1-2 weeks, if it was allowed to trade in the unfavorable phase of the market, It was a series of losing trades. On average, one trading operation per day was obtained, which, given the possible large interruptions in trading, again would not be very clear for demonstration on a demo account. In this regard, I did not begin to complete this trading system.

Debugging these two trading systems further strengthened my idea that a trading system should be self-learning and, in the course of work, rebuild its parameters to a changing market situation.

I decided to make another trading system to demonstrate the capabilities of indicators, but based on a neural network with self-learning elements, and prompt tracking of a changing market situation. Yesterday I wrote this trading system, but its debugging will take some time, and having spent time debugging the first two trading systems, I took too long to prepare for the presentation of indicators. Therefore, I decided, without waiting for the completion of the debugging of the trading system, to make a presentation of the indicators, and when the trading system is ready, to demonstrate its operation.

____________________________________________


Those wishing to buy indicators apply to the address:

piligrimm _ fx {dog} mail . en


If you have any questions about the operation of indicators, suggestions for their improvement, feedback on their use, write in this thread.

This topic was created with the aim of presenting the programs I develop, I hope it will be voluminous and it will contain many developments useful for trading. In this regard, we kindly request: do not load this topic with irrelevant discussions , so that potential buyers, for whom it is intended, do not lose answers to their questions among irrelevant discussions.

____________________________________________

Files:
 

Why are you selling if it's really cool?

Looking at the pictures - didn't understand anything.....

 
LeoV:

Why are you selling if it's really cool?

Looking at the pictures - I don't understand anything.....

What's not clear, can you be more specific?

 
Piligrimm:

What is not clear, can you be more specific?

Well, for example, on "Kristi_GrafWav" there are a lot of different coloured lines. What do they all mean? What are the rules for entering and exiting?

 
LeoV:
Piligrimm:

What is not clear, can you be more specific?

Well, for example, on "Kristi_GrafWav" there are a lot of different coloured lines. What do they all mean? What are the rules for entering and exiting?

Each line is obtained as a result of filtering by a group of wavelet transforms with different parameters, one or more tools specified in the indicator description. Entry and exit rules are selected according to the strategy you create, in the simplest case it can be crossing of single lines with each other or change of movement direction points.

 
Piligrimm:

I wanted to give a demonstration of how the indicators work on a demo account using the investor password to access the account for everyone, unfortunately something is not working. I have never used investor password before, but now I tried to log in to the account from another computer, but the corresponding windows with the instruments and indicators in use are not visible. Can you tell me how to do it correctly or is it only possible to trade with Expert Advisors via investor password?

Strange to hear such a thing from the developer :(

The indicator (and the Expert Advisor too) works and draws only in the terminal and PC in which it is installed.

But how else?

Through the investment entry you can only see the level of order open, stop and take. And the trades can be placed on the chart.

 
Solid experience in neural networks etc. is of course very good, but questions arise. Take the first picture. Let's point a finger at any point and ask ourselves a trivial question: where to, up or down? Besides, you should take into consideration that it is M1 and there is not much time to think it over. And there is also a question of the exit. And again, if you trade on the daily chart, it is not a problem, but on М1, and even with taking into account the spread, you may catch a loss.
 
The other drawings show H1, but that doesn't make it any easier. To be honest they are no more informative for me than 10 MA on one chart and tweaked Envelopes on the other. All this is of course my IMHO (!) and I'm by no means trying to offend the author. Just if the idea is good, it will stand up to healthy criticism.
 
sayfuji:
Solid experience in neural networks etc. is of course very good, but questions arise. Let's take the first picture. Let's press the finger to any point and ask a trivial question: up or down? Besides, you should take into consideration that it is M1 and there is not much time to think it over. And there is also a question of the exit. And again, if you trade on the daily chart, it is not a problem, but on М1, and even with taking into account the spread, you may catch a loss.

I think the jurik is much faster...... And the smoothness is just as good.... Have you tried comparing it to jurik?

 
What is the pink line in the Indicator Channel indicator?
 
Piligrimm:

A detailed manual is included, with which even the inexperienced user of Matlab can easily do everything. Matlab can be downloaded from: http://torrents. ru/forum/ .

....

Cost of indicator "Kristi_GrafWav": $85 US.

Rare crap.

First recommends stealing other people's software and then tries to sell your own.