As, an afterthought, please note âMost profit is not always the best ideaâ. With reference to the earlier post, wherein the results of the system THREEMACROSSOVERS and its optimization were posted. Here, the rest of the experiment has been posted. We used the tool âGA Optimizerâ for the rest of the experiment. It is âGenetic optimizationâ, an Add-in of TradersStudio and is used for optimization of the results. Along with this tool, an Excel file (TSData.xla) is provided. The steps done for the rest of the experiment are as follows: We integrated the Excel file (TSData.xla) into Ms-Excel via the menu âTools->Add-Insâ. You can import the optimization results of the system THREEMACROSSOVERS, into Excel via this tool. After this, a new menu âProcess Dataâ comes, which shall be used for analyzing of the results. Before we analyze the results, as a precaution, a copy of the excel file (optim_MA_Crossovers.xls) was created (Copy of optim_MA_Crossovers.xls). We need to create 3 TripleMACrossovers as follows. Please do note that Param1 is SLen, Param2 is Mlen and Param3 is LLen. SLen < MLen and then, SLen < LLen MLen < LLen and then, SLen < LLen SLen < LLen and then, SLen < MLen Now, for the first one, SLen < MLen and then, SLen < LLen Open âCopy of optim_MA_Crossovers.xlsâ in excel. We need to process the excel file data, this is done via the menu âProcess Data->Split Dataâ This creates a process data worksheet. Here, the idea is to separate the cells from comments into their own cells for further processing. Now, here we create a 3D-Optimization chart with parameters and filtered data via the menu âProcess Data->3D-Chartâ. This has to be done carefully and in such a manner that SLen < MLen. Therefore the steps here are: a) A dialog box comes wherein x, y & z axis needs to be given. We gave x value -> Param1, y -> Param2 & z -> Net Profit. b) Now, the âFilter Dataâ button is pressed wherein we give the condition that SLen < MLen. Thus, we fill up the boxes as Column 1 (Param1), Column 2 (Param2) & Condition (<) c) A chart is created. We now save the excel file as âp1&2_optim_MA_Crossovers.xlsâ, in order to keep the original file safe. Next, is âProcess Data->Filter Dataâ, here we mention the next condition that SLen < LLen i.e. Param1 < Param3. Lastly, we do âProcess Data->Validateâ, in order to analyze the result and check its robustness. The results for this has been posted as an excel file (p1&2_optim_MA_Crossovers.xls). Now, for the second one, MLen < LLen and then, SLen < LLen Steps 1 & 2 same as in the first one. Now, here we create a 3D-Optimization chart with parameters and filtered data via the menu âProcess Data->3D-Chartâ. This has to be done carefully and in such a manner that MLen < LLen. Therefore the steps here are: a) A dialog box comes wherein x, y & z axis needs to be given. We gave x value -> Param2, y -> Param3 & z -> Net Profit. b) Now, the âFilter Dataâ button is pressed wherein we give the condition that MLen < LLen. Thus, we fill up the boxes as Column 1 (Param2), Column 2 (Param3) & Condition (<) c) A chart is created. We now save the excel file as âp2&3_optim_MA_Crossovers.xlsâ, in order to keep the original file safe. Next, is âProcess Data->Filter Dataâ , here we mention the next condition that SLen < LLen i.e. Param1 < Param3. Lastly, we do âProcess Data->Validateâ , in order to analyze the result and check its robustness. The results for this has been posted as an excel file (p2&3_optim_MA_Crossovers.xls). Now, for the third one, SLen < LLen and then, SLen < MLen Steps 1 & 2 same as in the first one. Now, here we create a 3D-Optimization chart with parameters and filtered data via the menu âProcess Data->3D-Chartâ . This has to be done carefully and in such a manner that SLen < LLen. Therefore the steps here are: a) A dialog box comes wherein x, y & z axis needs to be given. We gave x value -> Param1, y -> Param3 & z -> Net Profit. b) Now, the âFilter Dataâ button is pressed wherein we give the condition that SLen < LLen. Thus, we fill up the boxes as Column 1 (Param1), Column 2 (Param3) & Condition (<) c) A chart is created. We now save the excel file as âp1&3_optim_MA_Crossovers.xlsâ, in order to keep the original file safe. Next, is âProcess Data->Filter Dataâ , here we mention the next condition that SLen < MLen i.e. Param1 < Param2. Lastly, we do âProcess Data->Validateâ , in order to analyze the result and check its robustness. The results for this has been posted as an excel file (p1&3_optim_MA_Crossovers.xls). All the three above excel files plus the copy of the optimum excel file have been posted as âTripleMACrossover.zipâ for your kind reference. Please do not hesistate to ask any queries, if you have.
How is the parameter âmaxbarâ important? Do we have to take results from such a old date as 1980? What will happen if we take another different date?
TraderSystem, you have asked a good question. I am not sure of the answer, shall need to think it over, or still better I shall ask Mr. Murray. Mr. Murray, please answer the above query, thanks.
Attached is a modified ThreeMACrossover code[ Here, we shall add conditions to the part with the buy /sell rules to filter trades i.e. we shall add filters like ADX(14)>30 or RSI(CLose,10,0)>80 for a buy, so that you are not already overbought. We shall continue this in the next post.
What kind of filters, would you like me to apply in order to improve my triple MA system? I do have some ideas, however, I would like to have your inputs and more ideas. Thanks!
QUOTE]Quote from Gyles: As, an afterthought, please note ââ¬ÅMost profit is not always the best ideaââ¬Â. With reference to the earlier post, wherein the results of the system THREEMACROSSOVERS and its optimization were posted. Here, the rest of the experiment has been posted. We used the tool ââ¬ÅGA Optimizerââ¬Â for the rest of the experiment. It is ââ¬ÅGenetic optimizationââ¬Â, an Add-in of TradersStudio and is used for optimization of the results. Along with this tool, an Excel file (TSData.xla) is provided. ............................. Please do not hesistate to ask any queries, if you have. [/QUOTE] I am confused, did you use "GA Optimizer" or "3-D Optimizer"? The excel file "TSData.xla" is with 3-D one.
Thanks TraderSystem, for pointing out the error. You are correct, for this part I used ââ¬Å3-D Optimizerââ¬Â and not ââ¬ÅGA Optimizerââ¬Â. Sincere apologies for the confusion These two are different as ââ¬Å3D optimizerââ¬Â analyzes optimization grid while ââ¬ÅGA Optimizerââ¬Â runs a smart optimization. In fact, I am using both these add-ins for this experiment: The ââ¬Å3D optimizerââ¬Â with the optimization grid. The ââ¬ÅGenetic optimizerââ¬Â in order to do major optimization test for filters later on.
I am posting the reply again as the above answer looks garbled up and confusing. Thanks TraderSystem, for pointing out the error. You are correct, for this part I used "3-D Optimizer" and not "GA Optimizer". Sincere apologies for the confusion. These two are different as "3D optimizer" analyzes optimization grid while "GA Optimizer" runs a smart optimization. In fact, I am using both these add-ins for this experiment: The "3D optimizer" with the optimization grid. The "Genetic optimizer" in order to do major optimization test for filters later on.
Is it possible to post more details, regarding the experiment, something like they do in manuals, explaining steps with images, so it is easier to understand?