Track trigger timing

[David McFarlane]

Evelina,

Mich pretty much covered it.  I will add only the following:

For debugging purposes, you can sprinkle the code with MsgBox or 
Debug.Print statements so that you can trace execution flow and see 
if the code really goes where you think it does.  MsgBox will mess up 
the program timing, but will show at run time where your code is, and 
will work outside of E-Studio; Debug.Print requires running from 
E-Studio, but offers some other niceties.

I would not use a c.Log anywhere except in exceptional circumstances 
(e.g., a loop within a Procedure where I want to add extra log 
lines).  The Procedure already supplies a c.Log where needed, each 
extra c.Log that you add also adds another line to the log, which is 
rarely what you want.  More to the point, simply using c.SetAttrib 
already does everything needed to added an attribute to the log, 
since EP automatically addes the required c.Log (don't take my word 
for that, look at the generated code just before End Sub of your Procedure).

The PulseOn/Off need not be "attached" to any object in the program 
(and as Mich pointed out, you don't really need the variables at 
all).  As described it looks like you are doing everything correctly, 
so we must be missing something.  Which brings us back to the general 
debugging tips offered by Mich & me to see if the program flow really 
goes where you think it does.

Good luck,
-- David McFarlane, Professional Faultfinder


Michiel Spape wrote:
>Hiya,
>What you're saying doesn't sound like anything E-Prime should omit. 
>The only reason I can see your data is not being saved is that for 
>some reason, these lines are skipped. There can be many reasons for 
>this (say, a button-press causes a goto to fire), but what you would 
>need to find out is whether you are actually logging anything. That 
>is, just move the inline thingy forward/backwards along the 
>procedure and see whether anything is logged. I'd also recommend 
>shortening the line to
>c.SetAttrib "PulseOnset", clock.read
>so that you don't even have to declare any variable.
>
>One question though: is _nothing_ being saved, or is PulseOnset 
>found somewhere in your edat file, but with nothing in there? If the 
>former, then the line is presumably skipped (as i had assumed based 
>on what you write). If the latter, then a number of other things may 
>be going on (e.g. it's saved under a sub-trial level thus easily overlooked).
>
>One final thought: I'd first get everything running, THEN start 
>thinking about adding TMS triggers and whatnot. It's far easier to 
>add advanced bits to the working basics than vice versa. Can you, 
>for instance, get a Stroop experiment (or any of the tutorial 
>experiments) running with a c.SetAttrib "MyExtraLoggedData", clock.read added?
>cheers,
>Mich
>
>________________________________________
>From: e-prime at googlegroups.com [e-prime at googlegroups.com] On Behalf 
>Of Evelina Tapia [evelina at illinois.edu]
>Sent: 04 January 2011 18:13
>To: E-Prime
>Subject: Track trigger timing
>
>Hi,
>
>I am creating an experiment where a TMS is triggered at various points
>during a trial sequence (where several stimuli are presented). I have
>InLine code at the very beginning of the experiment that randomizes
>the onset of the trigger. What I want to be able to do is to timestamp
>the execution of the trigger -- in other words, I want a timestamp
>when the line of code in InLine for trigger TMS is begun/completed and
>I want to log that data together with other information in the
>final .edat file. For that purpose I have declared 2 global variables
>"PulseOn" and "PulseOff" under the User tab in the Script window. In
>my InLine code right after trigger is executed (and I know that the
>trigger works just as it should) I have added
>
>PulseOn = Clock.Read
>c.SetAttrib "PulseOnset", PulseOn
>
>and similar code is placed after the port is turned off for trigger
>offset time stamp. However, when I run this code, the variables
>PulseOnset/Offset are not saved in the data file. I have also tried to
>use c.Log in this InLine but then only one line of data (one trial) is
>saved and although trigger onset/offset are marked, there is no more
>any data about the objects presented during the trial.
>
>In the documentation I have found that "a variable must be set as an
>attribute of the context object in order to be logged in the data
>file" so I assume that the issue here is that PulseOn/Off aren't
>really "attached" to any object in the experiment... Is there a way
>around that? Or another way to log the onset/offset of the trigger?
>
>THANK YOU!
>
>Evelina

-- 
You received this message because you are subscribed to the Google Groups "E-Prime" group.
To post to this group, send email to e-prime at googlegroups.com.
To unsubscribe from this group, send email to e-prime+unsubscribe at googlegroups.com.
For more options, visit this group at http://groups.google.com/group/e-prime?hl=en.