I was always interested in zig zag analysis and was about to write a generic zigzag analyzing tool anyways (a zig zag which computes the best fitting zig zag for a period and projects it on the next session). Is there someone coding this stuff for you? By skimming your start post you have a lot of ideas which are actually kind of simple to implement but since there are so many ideas it will take a few hours to get done with it. If you are in need of someone please send me a message next monday evening than I have time to look into it.
Drill yourself quickly into pairs you prefer: Zig zag statistics
- #42
- Edited 5:11am Jul 13, 2013 4:35am | Edited 5:11am
- | Joined Mar 2011 | Status: I have a member | 674 Posts
I was always interested in zig zag analysis and was about to write a generic zigzag analyzing tool anyways (a zig zag which computes the best fitting zig zag for a period and projects it on the next session). Is there someone coding this stuff for you? By skimming your start post you have a lot of ideas which are actually kind of simple to implement but since there are so many ideas it will take a few hours to get done with it. If you are in need of someone please send me a message next monday evening than I have time to look into it.
At the moment I dont have anyone coding it so It would be great to cooperate with this. Your idea sounds interesting. Would you care to share more of your thoughts here or by PM?Monday evening is fine. Im available after around 15:00 PM gmt.
- #43
- Jul 13, 2013 5:33am Jul 13, 2013 5:33am
And you should be sure about what you want 
. I prefer statistic indicators everyday over trading signal indicators.
A year ago I wrote an indicator which sampled the volume of a few years worth of 30 minute bar data to give me a picture about what the average volume on a day looks like. I got a different bell shape curve for every specific day of the week and you can see how the volume comes close to the approximation every day. Once the volume is far of you know something is wrong. I like math and I had to sit in a statistics lecture last year so I am into bell curves Brownian movements etc. Statistics is quite fascinating once you understand what is behind it.
Just some thoughts about how to transfer data:
1: Show the final data on the chart and manually input them into Excel.
2: File write function. You have to open an css file manually and got all you data there.
3: There is a way to import data directly into a excel spreadsheet using DDE (dynamic data exchange). Once the value on the chart changes the value in the cell in excel changes as well. (live update).
I have just worked with DDE once and I know that you can get the Open Close High Low values of a bar but I am not sure if you can send custom values like we want to. A workaround would be to create a synthetic currency pair and feed with with our own data and export those values using DDE. (This goes a little bit deeper into mql4).
4: You can export data from mql4 using pipes/streams. I am not sure how Excel handles those though and for this we need to include the windows api. Thats like a whole different level of programming again.
. I prefer statistic indicators everyday over trading signal indicators.A year ago I wrote an indicator which sampled the volume of a few years worth of 30 minute bar data to give me a picture about what the average volume on a day looks like. I got a different bell shape curve for every specific day of the week and you can see how the volume comes close to the approximation every day. Once the volume is far of you know something is wrong
. I like math and I had to sit in a statistics lecture last year so I am into bell curves Brownian movements etc. Statistics is quite fascinating once you understand what is behind it.Just some thoughts about how to transfer data:
1: Show the final data on the chart and manually input them into Excel.
2: File write function. You have to open an css file manually and got all you data there.
3: There is a way to import data directly into a excel spreadsheet using DDE (dynamic data exchange). Once the value on the chart changes the value in the cell in excel changes as well. (live update).
I have just worked with DDE once and I know that you can get the Open Close High Low values of a bar but I am not sure if you can send custom values like we want to. A workaround would be to create a synthetic currency pair and feed with with our own data and export those values using DDE. (This goes a little bit deeper into mql4
).4: You can export data from mql4 using pipes/streams. I am not sure how Excel handles those though and for this we need to include the windows api. Thats like a whole different level of programming again.
- #44
- Jul 13, 2013 12:37pm Jul 13, 2013 12:37pm
- | Joined Mar 2011 | Status: I have a member | 674 Posts
And you should be sure about what you want
. What I want is more or less described in the first post, but the development journey itself produces some aha moments also. I kinda see this as a two way trip. On the other hand the end result could be an indicator which shows where we are at in a particular moment in time. On the other hand there is the longer term statistical modeling of the pair, which is of interest also. A real time data collector would be a significant help there. As I trade mostly during NY session, it is of particular interest to me. So the possibility to analyse sessions statistically is important to me.
Maybe those two views can be at least partly combined like if the "statistical dashboard" of the indicator could show where we are at statisticswise. Eg. if your trading on M1 or M5 it would be nice to see how far have the ZZlegs gone in longer TFs and what are the probabilities for them to go further..... Just some thoughts
.
- #45
- Jul 13, 2013 12:47pm Jul 13, 2013 12:47pm
. if your trading on M1 or M5 it would be nice to see how far have the ZZlegs gone in longer TFs and what are the probabilities for them to go further..... Just some thoughts
. Well dang I should get back to studying. Talk to you later.
{quote} That sounds doable and should work similar like my fibonacci probability analysis tool
No harm to have them others also, but maybe adjustable. I have previously been studying fib use also on lucas numbers & gann angles & such. As the market is kinda fractal, other important mathematical numbers seem to work also in addition to fibs, but fib 50 is the most important one I think. But I think too that you should study as its a bit late to do that on monday night
See ya.
- #48
- Jul 13, 2013 1:23pm Jul 13, 2013 1:23pm
I remember having issues getting all the needed historic data last time I did an extensive analysis. Are you able to download enough historic data from the metaquotes server directly? It would be great if we work on the same sets of data would you mind uploading your historic data? That would make this much easier . By the way thank you for your work and dedication.
. By the way thank you for your work and dedication.
I remember having issues getting all the needed historic data last time I did an extensive analysis. Are you able to download enough historic data from the metaquotes server directly? It would be great if we work on the same sets of data would you mind uploading your historic data? That would make this much easier
I also think that the data should be analysed according to different markets. Average true range is a good tool to divide different market conditions as the swing sizes are different when it is smaller than if its bigger as I illustrated on some previous post. I will get back into this and show it in a bit more detail laters.
And thanks for thanking
. Nothing comes without work and dedication. Especially in forex.....
- #50
- Jul 13, 2013 2:04pm Jul 13, 2013 2:04pm
{quote} I can try, but I think there may be other issues in the data also as it may differ on different brokers. Extensive amount of data also easily jams up my computer
). Well lets not get ahead and start with the most basic stuff huh. Adding more functionality later is always possible. As long as you don't let metatrader connect to the internet you can load historic data into it from any broker you want. Thats why I ask for your data so we work with the same stuff.
- #51
- Edited 4:33am Jul 14, 2013 3:01am | Edited 4:33am
- | Joined Mar 2011 | Status: I have a member | 674 Posts
{quote} How about a wSR value (weighted swing range based on the ATR value
On the statistical analysis part of the project I was also thinking of cohort analysis of a kind based on the levels on ATR. To create different sets of "subdata" on each cohort.
About the data. Do you know if theres any free external cloud server, etc. where we could put the data and use same the data together? I think it could be the simplest solution to data management problems. The data earlier than 1990īs seems to be more or less inadequate.
Edit: Sample pic of cohort distributions enclosed. The problem with cohorts is that they have to be manually defined.
Edit: The formula for cohort definition to be enclosed to the excel on the first post of this thread in case someone likes to play with them:
=IF(L2<0,8;"1";IF(L2<1,3;"2";IF(L2<1,8;"3";IF(L2<2,2;"4";IF(L2>=2,2;"5")))))
- #52
- Jul 14, 2013 11:03am Jul 14, 2013 11:03am
- | Joined Mar 2011 | Status: I have a member | 674 Posts
{quote} How about a wSR value (weighted swing range based on the ATR value
