Realtime Swings.efs

[ACM]

jrbarr
FWIW I tried running both and am not getting any syntax errors on either one (version 7.6 build 636a)
Alex

[jrbarr]

Thanks Alex for testing this. Do you have any idea why I'm receiving these error messages or what they signify?

At any rate, I'll download Build 636a, run these scripts on that version, and see if the error messages continue.

jrbarr

[ACM]

jrbarr
I can't think of any reason as to why you would be getting a syntax error.
BTW are you getting the same error message for both efs?
If not what is the message for the revised version?
Alex

[jrbarr]

Hi Alex,

I downloaded Build 636a, but the error messages persist. I only tried "RealTimeSwings.efs (Version 1.1 – Updated 4/8/2004)" today, hoping it wouldn't produce the error message, so I don't have a history of use with it. I've been using RealTimeSwings2.efs for about 10 days, and the problem has been fairly persistent. Typically, upon receiving the error message, once I "reload" the script it will work fine.

I've tried isolating any constant that might trigger the error message, such as intraday vs. daily charts, continuous vs. quarterly contract, etc., and my sense is that is appears most often on smaller timeframe intraday minute-based charts, vs. longer intraday minute-based or daily charts. I will continue to monitor this.

The only other constant that I can find is in the code. The term, "SwingPtsDev", is only used twice in the script, and it is on both lines 323 and 328, which are the only lines referenced in the error messages. From what I gather, this section of the code refers to the currently developing swing line.

Thanks,

jrbarr

[gspiker]

What symbol,interval are you using? Also what does the Time Template look like, and what Time Zone are you in?

Param #3 looks to be a bar offset.. (unless the error message counts by 0...not to clear on that but it would be easy to test and I will once I have more info).

Garth

[jrbarr]

Garth,

Try NQ M4 @ 5min or ES M4 @ 2 min, as examples of symbols and intervals that produce the error message. Time 6:00 to 13:00 in the Pacific Time Zone.

Thank you for your willingness to test this. It is a great script.

jrbarr

[gspiker]

Well I can't reproduce it either, given the parameters you defined.
Does this happen in RT mode only, or does just loading the EFS do it?

I did trying running in replay mode (not the same as RT obviously), but got no error.

I did verify that param 3 is bar offset for the second point in the line. Not sure that helps, but at least we know which param is giving the error.

I also noticed that the line numbers you give don't line up to my line numbers...in other words I don't have drawLineRelative() commands on the same lines you do.

If the errors happen on formula load my suggestion:

1) Download RealTimeSwings.efs again.
2) If the problem still exists - add a debug line in front of the drawLineRelative() commands...

debugPrintln(" x1b = " + x1b);

and see what the last message is right before you bomb out. Then let us know.

If it only happens in RT...then it will have to wait for Monday for me (or Jason) to debug it.

Garth

[jrbarr]

Hi Garth,

I am very grateful for all your help and insights.

1) Just loading any layout containing this script will trigger the error message. It does not need to be in realtime.

2) I'm glad you could pinpoint the source of the error message.

3) I also can't account for the difference in line numbers. Per your suggestion, I downloaded the script again and placed it in a different folder. The problem persisted, although it seems less frequently. I also added the debug line, which tripped one time and produced a whole new set of error messages.

I'll continue to monitor.

Appreciatively,

jrbarr

[gspiker]

jrbarr,

You should have seen the debug message more than once. Did you add it before both of the DrawLineRelative() commands?

What was the value of x1b when the error happened?

Garth

[jrbarr]

Yes, Garth, I added the debug message before both of the lines. In fact it has now consistently appeared more than once.

Here is a graphic with its values.

Thanks,

jrbarr

[eSignal_JasonK]

Hello jrbarr,

It looks to me like the formula you are currently using is the old version. Line 328 in the most recent version is a blank line and 323 is: "aSwingsIndex[0] = x1b;".

Open the formula, RealTimeSwings.efs (not RealTimeSwings2.efs), in the editor and tell me what you see on line 4.

[jrbarr]

Hi Jason,

Both RealTimeSwings.efs and RealTimeSwings2.efs list Version 1.1 on Line 4. Both scripts again list the date 3/24/2004 under Notes on Line 27.

However, the difference that you and Garth both noted concerning the anomalies in the line placement when comparing the two scripts is the only tangible clue. Line 323 in the new version is indeed, "aSwingsIndex[0] = x1b;".

I will remove the old version from all my charts and install the new version and see if that solves the problem, and I will report my findings.

Thanks very much,

jrbarr

[jrbarr]

Hi Jason,

I've been using the new version of RealTimeSwings.efs for several days, and it has worked perfectly with no glitches. Thank you, Alex, and Garth for the time and effort you each put forth to solving this.

jrbarr

[Anson]

Hello!

Yes, I agree it's a great job.

[BakedWafer]

Could someone use this in an autotrading strategy?