Jump to content

shawnh

⭐ V.I.P.
  • Posts

    165
  • Joined

  • Last visited

  • Days Won

    2

Posts posted by shawnh

  1. Hello guys, I've been playing around with Advanced GET 9.1 EOD Dashboard version, trying to get it working with intraday data. After some rough patches and trial and error, I think I have things working fairly smoothly. I'd like to share what I've learned with others who may be in the same boat.

    I use the technique of getting real-time data from an MT4 demo... this method is explained in another thread. Since AG EOD can't accommodate the "time" field of a bar, the MT4 intraday data has to be converted to "pseudo" daily data via an MT4 indicator placed on the intraday charts. Works well, but I lamented the fact that once I viewed the data charted in AG, I could no longer see the correct date/time of a bar. So I did some modifcations of the original MT4 indicator to put the proper date/time stamp of each bar in the VOLUME field instead. So if you don't need the VOLUME in any of your analysis, this is for you.

    Here is the modified indicator for MT4:

     

    https://anonfile.com/f040Bfbebe/AgetDataDatestamp.rar

     

    When placed on a MT4 chart, it will output an ascii file to your MT4 installation folder in the "Files" subfolder. Now, when you open up a chart in AG using this data (and select the "Volume" study), you see this:

     

    https://anonfile.com/gf40Bcb9bf/ShowDatestamp.jpg

     

    Way down there at the lower left will be the proper DateTime stamp of the bar your crosshair is placed over... instead of the Volume figure. I like to grab the upper border of the Volume study pane and drag it down so just the DateTime value shows. It seems to work pretty good, but alas, it suffers the same fate as all studies placed on AG charts once you close down AG and restart it - they disappear or are disabled. You'll notice that the Volume study is still there at the bottom, but scrolling the crosshairs won't change the value. Quick fix: just right click over the word VOLUME. click "Properties", then "Apply" and OK. It will come back to life. Easier than deleting the whole study and putting it back on.

     

    At first I had AG on my Vista desktop and I struggled to try and get it working half-decently. Constant freeze-ups, very slow page loading, etc. I tried it instead on my older and slower XP laptop and it seems to work *much* better. And here is probably the most important thing I discovered to make closing/re-opening AG work quick and smooth - the power of the PAGE|NEW PAGE button! When you're in AG and working on a page of charts (I have 6 charts open in the one page I have - 3 AAPL and 3 SPY) and ready to exit and re-open, follow these exact steps (works for me!):

     

    - Save your page using PAGE|SAVE PAGE (... or Save as)

    - click PAGE|NEW PAGE to clear out the screen (*always* do this just before closing up AG!)

    - Close up AG, when it asks "Save the current page", click NO. (you've already saved it)

     

    To re-open AG, I have a small batch file already prepared that suppresses the loading of the eSignal module. This seems to save a lot of headaches and freeze-ups. Put the following in your batch file (ie. a .BAT file, and just place it on your desktop):

     

    CD \

    CD Program Files\Advanced GET EOD

    start AdvancedGETEOD.exe /d

     

    That's it. Launch that... AG will fire right up quickly to an empty screen, thanks to you doing the NEW PAGE thing prior to your last exiting. Go to PAGE|LOAD PAGE and load up your page and those charts should come right up pretty quickly.

     

    Another small error I got en route to figuring stuff out was the one below:

     

    https://anonfile.com/e749B7b5b3/AGerror.jpg

     

    To solve that, I would open up my page of charts, then do a SAVE AS, saving them under a new page name. Then deleting the old page. That took care of it.

     

    So I find now, finally, this works pretty good. Not perfect, but it's quite usable. Still remaining is the annoying thing of why the studies disappear off the charts everytime you shut down/restart. The drawn objects however - like the MOB and TJ's Ellipse remain.

     

    Hope this helped someone...

     

    Shawn

  2. Thank you wcicom for the advice on trading Gartley patterns - I look forward to dipping my toes into them. Very interesting that you helped with the MotiveWave project - it is a nice software. Perhaps you would have some insight about something I noticed - sometimes it has a very confusing habit of "bunching up" all the wave forecasts together at the end-of-day marker. Please see screenshots below.

     

    https://anonfile.com/saJ457b4ba/MWforecast1.jpg

    https://anonfile.com/r1Ja59baba/MWforecast2.jpg

     

    Maybe there is some setting that rectifies this...?

     

    Thanks!

    Shawn

  3. Thanks guys! I have installed and tried Elwave v9.2 and it does work with the IB datafeed. The warning message on the right saying "illegal copy, results will be distorted" is a little unsettling though. Elwave seems to be slow to respond when it is connected to the datafeed also.

     

    Guys, do you know if there is a version of Advanced Get that will read Interactive Brokers data feed?

     

    Ciao

    Shawn

  4. Yes. on daily time frame those Gartley patterns can be massive. Who can wait that long?? Do you put much weight in the "scoring" system that the scanner of Motivewave uses... ie. 1, 2, or 3 green dots. Do you tend to go for only the 3-green-dot recommendations, or it really doesn't make much difference?

     

    I have run into another puzzling issue with MotiveWave also - when applying an Elliott Wave "Auto-analyze" to a chart of 5-minute bars that I have set to go back to Jan 15/2016, it will give the error message "Unable to auto-analyze. Not enough data is available". I assumed that my brokers datafeed (InteractiveBrokers) would not supply 5-min bars to go back that far, so I got my hands on an ascii file of 5-min bar history going from Jan 1/2016 to June 30/2016... plenty (I thought) to fill in the gap. I imported it in correctly (at least it seemed so) and ran the Auto-ANalyzer again but it gave the same message. Have you ever experienced any such problem with historical data?

     

    Thank you so much for the assistance Forexonly, I appreciate it. A Happy easter to you and your family!

     

    Cheers

    Shawn

  5. Thank you for the tips Forexonly! I did look a little at the scanner and Gartley patterns. I tried a scan on some daily data charts. The patterns produced however were quite long-term... it would take some months to play out. I think I agree with you that hourly time frame is more reasonable and useful. I trade stock options and I think this will be useful to me to find good setups and trades lasting a few weeks or so. Just right.

    As I was testing this, it occured to me that a Gartley pattern is not much different than an Elliott Wave A-B-C structure. I think running an automatic elliott wave count on a chart that shows a good Gartley pattern, would be a very good confirmation tool, no? If the "D" point of the Gartley corresponded with the end of an elliot wave 2, I think that would be a very very good signal for catching the start of an impulsive wave 3. What do you think? Perhaps you are already doing this...

     

    You mentioned too "Take completion of bats and sharks". What about the other types of Gartley patterns such as Butterfly, Crab, Deep Crab, Cypher, etc. Do you find that they don't work as well?

     

    Thanks!

    Shawn

     

     

    Hi

    I do use the scanner and it is very effective in finding set ups. I have tried many different time frames and I was only able to make money with:

    1. 1 hour charts

    2. Take completion of bats and sharks ( reversals, or finding tops or bottoms)

    3. Forex only ( hence my handle)

    4. you can try it with NT indicators as well, it's the set up that works, not the tool ( MW)

    it's good for 70-80% prob., but with 2 or even 3 to 1 risk reward, it works; why? I have no idea!!

     

    Good luck.

  6. Guys, I have solved my problem! I dug around a little more in the MotiveWave docs and it is all explained here:

     

     

    http://www.motivewave.com/support/how_to/elliott_wave_study.htm

     

     

    So if you notice that your wave counts for different timeframes are "disagreeing", just look back in a daily or weekly history chart of your instrument (ie. SPY) and locate a very major high or low. Use that date (or better - a few days before) as your START DATE when doing an elliott wave "Auto Analyze". Make sure "Minimum Bars" and "Maximum bars" are greyed out and put a check mark next to "Use Date". I also like to select "Decompose All". This way seems to give good solid EW counts agreeing across all my timeframes. Unfortunately, InteractiveBrokers data will not retrieve 5 min bars all the way back to the major low of Jan 19/2016, so I had to use the "intermediate" major low on Nov 4/2016. It seems to be still okay. Thanks for the input everyone... I hope what I have described here helps someone who was in the same situation as I was!

     

    Mr. Forexonly - you said "What did like on MW is the patterns ( bat, shark, etc..) on the hourly chart.". I have yet to explore this part of MW... I'm not really familiar with Gartley patterns but I will look into it. Have you tried scanning (and trading) for those patterns using Daily charts with success also... or only hourly charts? The built-in scanner of MW seems really useful for this. When a Gartley pattern is found, does it also give a price target that might be reached?

     

    Thanks

    Shawn

  7. Thank you for the feedback guys. I do understand that EW is not an exact science, with more than one different outcome... but still it's unsettling when you have so many time frames disagreeing. I trade 5min and 10min charts so I was hoping that MotiveWave would provide some halfway-useful counts for those shorter time frames, I do agree that the longer-term count should take priority, but I would still love to see more consistency and agreement among the counts of different timeframes.

     

    But doing some more digging in MotiveWave, I think I *may* have found what is contributing to this - scrolling back in the history of each timeframce chart, I see that the 30min and 1hr charts start their EW analysis/labeling on the major low of Nov 4/2016 (labelled as "0"). Whereas the 5min and 15min charts start their analysis on the high of March 15/2017. To me, this would cause a misleading short-term count, as the real recent major swing high/low was several days before that on March 1, no?! I'm no expert on Elliott Wave, but I have a feeling that's why the shorter timeframe counts are not in sync, in this particular case anyway.

     

    Maybe there is some setting to "force" MotiveWave to start its analysis on a certain date?? Thoughts anyone?

     

    Cheers

    Shawn

  8. Hello Group... I have installed MotiveWave v4.2.24 the other day (thank you Shenong). Has anyone else noticed that the Elliott Wave counts from different time frames sometimes do not "agree" with one another and are not in sync? It makes it hard to trust any analysis when one time frame says it is going in one direction, and another time frame says the opposite.

    Please have a look at the 2 screenshots attached of the intraday SPY action for today (Apr. 10). On the 5 minute chart, we have a count forecasting a decline down to around 231. But on the 1 hour chart EW forecast, it shows no move at all down to 231, but a rise up into the 240's. Very strange, no?

     

    Nice software MotiveWave - but such results make it kind of hard to trust! Has anyone else also experienced this.... maybe I am just doing something fundamentally wrong somewhere...

     

    https://anonfile.com/deA70db2b5/MW5min.jpg

    https://anonfile.com/e1A501bdbb/MW1hour.jpg

     

    Thanks!

    Shawn

  9. Thank you so much Sir! But I have to say - I do think the "warning" message show in v9.5 might have been legit! I do recall when I ran it (but only a few times), the results were a bit nonsensical sometimes! But I could be wrong..

     

    Anyway, thanks very much Banner.

     

    Cheers

    Shawn

  10. Thanks Sixer, but the EOD versions of Advanced Get just won't read the "Time" field of my data files, regardless of what the rest of the dats's format looks like... and the Real-Time versions can't seem to be directly configured to read in ascii files on disk (unless I want to go through all that TS 2000i Data Manager/Owndata ordeal, which is messy).

    So I don't see what difference really a formatting conversion tool would make.

     

    Thanks

    Shawn

  11. Well, I'll explain a little more what I'm trying to do Sixer. Right now I have Wave59 running and using my IB datafeed. In Wave59 there is a function "Output Chart to Ascii" that writes my S&P intraday charts O-H-L-C data for the last 4 or 5 days to a file on disk... and it updates it every bar (3 minutes). I also have MTPredictor running at the same time, to read this ascii file of S&P data, I simply have to manually refresh in MTP - works good. I am trying to get a version of Advanced GET that behave in this way too - so it could read this file of ascii S&P intraday data that Wave59 writes to disk.

    Advanced GET 9.1 EOD version is so close to be able to doing this, but it can't be configured to read data files with both "Date" and "Time" fields in it... only Date, O, H, L, C, V. At least I can't figure it out anyway....

     

    Thanks!

    Shawn

  12. Thank you Sixer, yes this is the exact version I have installed now (v1.1 for TS2000i, Build 120). I cannot see where to open an ascii file however. Clicking on "New Chart" brings up a window to let me select an existing symbol. but there are no symbols to choose from and I don't see anywhere how to add a new symbol, or specify a folder to read files from.

     

    Thanks!

    Shawn

     

    P.S: I have already tried v9.1 EOD Dashboard edition, but it would only read in End-of-Day data - no facility to specify the "time" field for intraday data :-(

×
×
  • Create New...