Lines of Code
[root]/trunk/task/trackers/Eyelink
| Author | Changes | Lines of Code | Lines per Change |
|---|---|---|---|
| Totals | 37 (100.0%) | 792 (100.0%) | 21.4 |
| dep | 18 (48.6%) | 400 (50.5%) | 22.2 |
| justin | 17 (45.9%) | 382 (48.2%) | 22.4 |
| eli | 2 (5.4%) | 10 (1.3%) | 5.0 |
Use the non-private callback.
2 lines of code changed in 1 file:
Renamed *EDF to EDF* to be consistent with other fns
[+ bugfix in merge of getTaskEyeTraces]
0 lines of code changed in 1 file:
Added MGL version and mglEyelink version to the preamble saved in
the edf file. this can be used to aid in parsing if changes or additions are
made in the future.
7 lines of code changed in 1 file:
*** empty log message ***
7 lines of code changed in 1 file:
added line to set the sensitivity of the event parser (defaults to 1, which is more sensitivity)
3 lines of code changed in 1 file:
Adding comment that callback names are required
6 lines of code changed in 1 file:
Adding ability to set number of calibration points by setting mglEyelinkParams
7 lines of code changed in 1 file:
converted all subfun isfile into a utility function
0 lines of code changed in 1 file:
A few more fixes so that the default (if you do not set any task to collectEyeData is to collect eye data for all tasks)
3 lines of code changed in 1 file:
Function to set/get eyelink parameters. There is corresponding code in initEyelinkEyeTracker that will read the parameters from this function, but it is not checked in yet until I can test it.
137 lines of code changed in 1 file:
Moving start eye link recording till after we calibrate the eye tracker
0 lines of code changed in 1 file:
initTask now always sets the variable collectEyeData so that we don't have to do a field check in the callbacks.
Also, if initTask has any task with a collectEyeData set to not false, then we turn of the myscreen collectEyeData which sets recording for the whole structure
27 lines of code changed in 5 files:
Adding taskID to messages for general saving of eye tracker data
16 lines of code changed in 2 files:
Default to starting eye tracker as soon as you call initEyelinkEyeTracker so that all data can be saved for all tasks
10 lines of code changed in 1 file:
Gives warning if you do not have eyelink compiled
6 lines of code changed in 1 file:
Commented out calibrate test message--it changes the mean luminance during autothresholding
6 lines of code changed in 1 file:
renamed mglPrivateEyelinkCalibration -> mglPrivateEyelinkSetup
2 lines of code changed in 2 files:
Number of minor changes here and there. The taskTemplateGazeContingent and calibrateEyeTracker sequence now display messages to get the experimenter started on what is going on.
170 lines of code changed in 3 files:
Merged revisions 455-463,465-466,470-471,476-488,490,492-493,496-497,500-505,508-509,516,521-522,524,527,529-534,536,539-542,544-562,564,624-625,632,644,646-649,652-653,656,658,672,674,682-683,686,692-693,700 via svnmerge from
http://gru.brain.riken.jp/svn/mgl/sandbox/eyelinkdev
................
r455 | dep | 2009-02-09 17:09:10 -0500 (Mon, 09 Feb 2009) | 8 lines
mglEyelink: a set of routines for accessing the Eyelink C API
mglPrivate funcitons are used for basic API calls
mgl functions are used for combining into useful groups (mgllib only)
There will need to be a mglEyeTracker set of funcitons for abstracting
away from specific eye trackers
There will need to be task based functions for data colleciton/gaze
contingent, trial calibration, etc.
................
r456 | dep | 2009-02-09 17:40:09 -0500 (Mon, 09 Feb 2009) | 3 lines
Added library include to the CFLAGS. This is machine dependent (e.g. could be in
the ~/Library/Frameworks or /Developer/Frameworks)
................
r457 | dep | 2009-02-09 17:42:15 -0500 (Mon, 09 Feb 2009) | 3 lines
DELETED mex binary files from this tree--they should not be updated after
builds and only source is really needed in the repository.
................
r458 | dep | 2009-02-09 17:46:35 -0500 (Mon, 09 Feb 2009) | 2 lines
Set the directories to ignore mex binaries.
................
r459 | dep | 2009-02-10 12:11:59 -0500 (Tue, 10 Feb 2009) | 2 lines
Developer information files for tracking TODOs and notes.
................
r460 | dep | 2009-02-10 15:54:15 -0500 (Tue, 10 Feb 2009) | 3 lines
Base calibration code copied from sdl plugin code example
Fixed: Added include for eyelink_core in mglEyelink/mexopts.sh
Fixed: Added frameworks to get mgl compiling on 10.4/ppc
................
r461 | dep | 2009-02-10 23:40:49 -0500 (Tue, 10 Feb 2009) | 3 lines
Bug in EyelinkOpen parsing; temp commented out
Calibration skeleton is working
................
r462 | dep | 2009-02-10 23:41:36 -0500 (Tue, 10 Feb 2009) | 2 lines
skeleton message passing
................
r463 | dep | 2009-02-11 10:26:27 -0500 (Wed, 11 Feb 2009) | 1 line
Send a command to the Eyelink
................
r465 | dep | 2009-02-11 11:57:22 -0500 (Wed, 11 Feb 2009) | 2 lines
mglMake: Changed the processing of the arguments to provide a more flexible and robust method for handling defines. The argument 'rebuild' can now be one of '1', 'rebuild', 'carbon', 'cocoa' or any define beginning with '__'. Additionally, any number of additional defines can be passed as further arguments.
mglMake: Also changed the processing of the mgl directory for sources to look in subdirectories beginning with 'mgl' and verified that 'mgl.h' actually exists in the main mgldir.
................
r466 | dep | 2009-02-11 11:59:27 -0500 (Wed, 11 Feb 2009) | 2 lines
fixed all the message sending private functions to take only a string--printf-like handling must be supported in .m files
renamed SetMode to GoOffline--this is the only command that is use to change mode prior to a series of commands (the others would be done in conjunction with beginning a trial, starting to record data etc.
................
r470 | dep | 2009-02-11 19:25:21 -0500 (Wed, 11 Feb 2009) | 3 lines
Bugfix: In mglMake when first parameter was passed as a define
Bugfix: added #ifndef protection to __eventtap__
................
r471 | dep | 2009-02-12 21:34:24 -0500 (Thu, 12 Feb 2009) | 1 line
Almost a complete working eyelink library.
................
r476 | dep | 2009-02-15 17:14:16 -0500 (Sun, 15 Feb 2009) | 2 lines
notes on textures and library
................
r477 | dep | 2009-02-15 17:15:52 -0500 (Sun, 15 Feb 2009) | 2 lines
removed mgllib notes to create a mgllib develoer.txt doc
................
r478 | dep | 2009-02-15 17:35:04 -0500 (Sun, 15 Feb 2009) | 2 lines
notes
................
r479 | dep | 2009-02-15 17:36:25 -0500 (Sun, 15 Feb 2009) | 2 lines
updated to provide options for specifying what data to be recorded
................
r480 | dep | 2009-02-15 17:36:57 -0500 (Sun, 15 Feb 2009) | 2 lines
more notes from psychtoolbox review
................
r481 | dep | 2009-02-15 17:38:09 -0500 (Sun, 15 Feb 2009) | 2 lines
display number for eye graphics
................
r482 | dep | 2009-02-15 19:04:28 -0500 (Sun, 15 Feb 2009) | 2 lines
place for developer notes for mgllib
................
r483 | dep | 2009-02-15 19:04:47 -0500 (Sun, 15 Feb 2009) | 2 lines
a wrapper for default ip
................
r484 | dep | 2009-02-15 19:05:10 -0500 (Sun, 15 Feb 2009) | 2 lines
fixed dereference bug
................
r485 | dep | 2009-02-15 19:05:38 -0500 (Sun, 15 Feb 2009) | 2 lines
updated to reflect new open wrapper
................
r486 | dep | 2009-02-15 19:05:54 -0500 (Sun, 15 Feb 2009) | 2 lines
merged in trunk line trace updates
................
r487 | dep | 2009-02-15 19:06:21 -0500 (Sun, 15 Feb 2009) | 2 lines
camera display work
................
r488 | dep | 2009-02-16 13:34:15 -0500 (Mon, 16 Feb 2009) | 1 line
added skeleton of eye camera code
................
r490 | dep | 2009-02-16 20:23:14 -0500 (Mon, 16 Feb 2009) | 2 lines
Compiles but crashes
................
r492 | dep | 2009-02-18 10:22:21 -0500 (Wed, 18 Feb 2009) | 2 lines
Fixed bug in int cast in open
basic calibration keycode framework
................
r493 | dep | 2009-02-18 10:22:54 -0500 (Wed, 18 Feb 2009) | 1 line
updated to allow general compile arguments (so it now requires -D for defines)
................
r496 | dep | 2009-02-18 18:38:25 -0500 (Wed, 18 Feb 2009) | 1 line
key code (untested)
................
r497 | dep | 2009-02-19 08:18:02 -0500 (Thu, 19 Feb 2009) | 1 line
fixed type bugs in crosshair code
................
r500 | dep | 2009-02-19 18:08:14 -0500 (Thu, 19 Feb 2009) | 1 line
Improved error message
................
r501 | dep | 2009-02-19 18:52:01 -0500 (Thu, 19 Feb 2009) | 2 lines
key events are being passed, but not all correctly
texture code is not completely failing...
................
r502 | dep | 2009-03-06 01:02:07 -0500 (Fri, 06 Mar 2009) | 2 lines
semi working version of the basic eye calibration display
................
r503 | dep | 2009-03-06 20:24:52 -0500 (Fri, 06 Mar 2009) | 2 lines
began re-factoring code to use c rather than matlab based gl texture calls
this will take a bit and could be the basis for a consoliation of the gl calls in the other texture functions
................
r504 | dep | 2009-03-07 19:06:52 -0500 (Sat, 07 Mar 2009) | 1 line
A basic version of the re-written code that uses c level gl calls based on the c mex functions that were previously used.
................
r505 | dep | 2009-03-08 15:53:55 -0400 (Sun, 08 Mar 2009) | 2 lines
Textures now blt to screen, but are the wrong shape and are are not updating correctly.
Blt does not re-bind the texture, this must be done externally, perhaps this suggests re-factoring...
................
r508 | dep | 2009-03-09 22:51:16 -0400 (Mon, 09 Mar 2009) | 3 lines
Basic work version of calibration utilities. The keyboard input is being passed correctly (except perhaps return/enter
not being different), the video works (but not the titles) but crashes/fails if you switch too many times. This is a
memory handling problem. Fixed (partially) broken carbon, non-listener code.
................
r509 | dep | 2009-03-10 14:03:54 -0400 (Tue, 10 Mar 2009) | 1 line
Working but listener is fucked.
................
r516 | dep | 2009-03-11 16:06:10 -0400 (Wed, 11 Mar 2009) | 1 line
Added unit test area and getkey test
................
r521 | dep | 2009-03-14 11:09:24 -0400 (Sat, 14 Mar 2009) | 2 lines
basic tracker outline
................
r522 | dep | 2009-03-14 13:22:03 -0400 (Sat, 14 Mar 2009) | 1 line
Check developer and global library for eyelink (should also check user)
................
r524 | dep | 2009-03-14 13:29:23 -0400 (Sat, 14 Mar 2009) | 1 line
Corrected developer framework location
................
r527 | dep | 2009-03-14 13:35:42 -0400 (Sat, 14 Mar 2009) | 1 line
decided to remove the /Developer include (use simlinks if desired) because it is more conforming with the Apple guidelines
................
r529 | dep | 2009-03-16 14:34:36 -0400 (Mon, 16 Mar 2009) | 4 lines
Began re-factoring c code into standard library format so to promote work on the mgl listener code
Eventually, the goals is to provide a general event stack which can be implemented for each platform.
Also, the goal will be to allow all mex functions to access the internal c calls used by other functions to promote
interoperability and flexibility of coding new functions. This will be demonstrated by re-factoring the texture code.
................
r530 | dep | 2009-03-16 16:53:37 -0400 (Mon, 16 Mar 2009) | 3 lines
Continuing saga of a calibration code gone wrong.
Restored separation of code, re-factoring will be too complex for now. it needs
to be done carefully and with coordination, so it is postponed.
................
r531 | dep | 2009-03-16 16:55:49 -0400 (Mon, 16 Mar 2009) | 3 lines
test change: attempt to flush the queue (which can build asyncronously) after a getkeyevent
call. this is the expected behavior, as the call should return the first event since the
last call (or you should get all keys).
................
r532 | dep | 2009-03-16 17:43:18 -0400 (Mon, 16 Mar 2009) | 1 line
Improvements/fixes to the task based callback system
................
r533 | dep | 2009-03-19 12:21:56 -0400 (Thu, 19 Mar 2009) | 2 lines
Attempt to use Carbon GetKeyEvent
................
r534 | dep | 2009-03-19 12:51:25 -0400 (Thu, 19 Mar 2009) | 1 line
Added init check and init set to determine if the tracker is initialized
................
r536 | dep | 2009-03-19 18:37:34 -0400 (Thu, 19 Mar 2009) | 9 lines
Merged revisions 535 via svnmerge from
http://yoyodyne.cns.nyu.edu/svn/mgl/trunk
........
r535 | dep | 2009-03-19 16:40:45 -0400 (Thu, 19 Mar 2009) | 2 lines
updates to (attempt) to fix listener/getkeys on carbon/10.4
........
................
r539 | dep | 2009-03-20 11:45:35 -0400 (Fri, 20 Mar 2009) | 1 line
a temporary test of Carbon Event Code
................
r540 | dep | 2009-03-20 12:38:22 -0400 (Fri, 20 Mar 2009) | 2 lines
test code
................
r541 | dep | 2009-03-20 13:12:55 -0400 (Fri, 20 Mar 2009) | 1 line
giving up on carbon events (non-threaded)
................
r542 | dep | 2009-03-20 18:19:27 -0400 (Fri, 20 Mar 2009) | 8 lines
Working mglEyelink.
This revision contains a working version of mglEyelink (the basic library calls) that
allows you to self calibrate, can return eye positions and appears (mostly) to not crash.
There is a lot of cleanup to be done in the code. Further, there MUST be a movement
towards a library based mgl with real make file and real architecture differentiation. Most
of what was tricky to make work was the eyelink calibration code (especially the eye drawing).
We still do not have working titles; we should report on state or return codes; etc.
Task integration is incomplete.
................
r544 | dep | 2009-03-20 18:39:08 -0400 (Fri, 20 Mar 2009) | 2 lines
get eye position in native or device coords
................
r545 | dep | 2009-03-21 13:51:57 -0400 (Sat, 21 Mar 2009) | 2 lines
tracker work
................
r546 | dep | 2009-03-21 14:05:55 -0400 (Sat, 21 Mar 2009) | 3 lines
use dirs for tracker code
eyelink callbacks
................
r547 | dep | 2009-03-21 15:45:52 -0400 (Sat, 21 Mar 2009) | 2 lines
baseline task support
fixed mglEyelinkOpen and dependents to allow dummy mode
................
r548 | dep | 2009-03-21 18:41:34 -0400 (Sat, 21 Mar 2009) | 2 lines
Working mglEyelink with tracking
................
r549 | dep | 2009-03-22 16:54:07 -0400 (Sun, 22 Mar 2009) | 3 lines
set physical params correctly,
fixed callbacks for trial/seg tags
................
r550 | dep | 2009-03-22 18:23:55 -0400 (Sun, 22 Mar 2009) | 2 lines
fixed callback code and changed default eye-coords
................
r551 | dep | 2009-03-24 12:09:22 -0400 (Tue, 24 Mar 2009) | 2 lines
Create a template gaze contingent task as a base for testing
bugfix: eyepos returned -y
................
r552 | dep | 2009-03-24 16:02:21 -0400 (Tue, 24 Mar 2009) | 2 lines
check to see if the save data field exists
................
r553 | dep | 2009-03-24 16:35:58 -0400 (Tue, 24 Mar 2009) | 2 lines
Added edf messages
................
r554 | dep | 2009-03-25 13:33:43 -0400 (Wed, 25 Mar 2009) | 4 lines
Added new callbacks,
Renamed recording functions,
Added task/phase specific recording spec
Integrated start/stop recording into the task structure
................
r555 | dep | 2009-03-25 17:02:54 -0400 (Wed, 25 Mar 2009) | 2 lines
Properly handle logical arrays
User friendly interface
................
r556 | dep | 2009-03-25 17:03:59 -0400 (Wed, 25 Mar 2009) | 1 line
removed system close call (unnecessary, i believe)
................
r557 | dep | 2009-03-25 17:04:18 -0400 (Wed, 25 Mar 2009) | 1 line
Checks to see if we are recording
................
r558 | dep | 2009-03-25 17:17:40 -0400 (Wed, 25 Mar 2009) | 2 lines
bugfix: integration of eye-tracker recording (mglEyelinkRecordStart) with task
................
r559 | dep | 2009-03-25 17:18:40 -0400 (Wed, 25 Mar 2009) | 2 lines
set the phase of a trial in the initTrial function (as seems appropriate)
alternatively it could be attached to the task struct directly (maybe more appropriate?)
................
r560 | dep | 2009-03-25 17:20:50 -0400 (Wed, 25 Mar 2009) | 1 line
removed debugging code
................
r561 | dep | 2009-03-26 11:22:44 -0400 (Thu, 26 Mar 2009) | 1 line
Use device coords by default
................
r562 | dep | 2009-03-26 17:59:20 -0400 (Thu, 26 Mar 2009) | 2 lines
bugfix: catch for next phase/task callback, check against length of task
................
r564 | dep | 2009-04-09 18:21:38 -0400 (Thu, 09 Apr 2009) | 2 lines
bugfix: StartRecording -> RecordingStart (and stop)
................
r624 | dep | 2009-08-27 16:16:18 -0400 (Thu, 27 Aug 2009) | 3 lines
Just moved the location of the task phase increment to be more rational
and allow the tracker callbacks to function
................
r625 | dep | 2009-08-27 17:00:23 -0400 (Thu, 27 Aug 2009) | 3 lines
Fixed eyetracker file save to not lose data on matlab crash
This should be improved, it's a bit of a hack.
................
r632 | dep | 2009-09-15 14:44:18 -0400 (Tue, 15 Sep 2009) | 2 lines
Fixed lack of file protection bug!
added test calibration and close
................
r644 | dep | 2009-10-28 11:29:29 -0400 (Wed, 28 Oct 2009) | 4 lines
fixes saving bugs for eyelink,
corrects recording check
rearranging getTaskParameters
................
r646 | dep | 2009-10-28 15:15:24 -0400 (Wed, 28 Oct 2009) | 2 lines
added ingore for mex on readDigPort
................
r647 | dep | 2009-10-28 16:44:22 -0400 (Wed, 28 Oct 2009) | 1 line
Fixed formatting and restored lost eyetracker initialization code
................
r648 | dep | 2009-11-04 16:39:58 -0500 (Wed, 04 Nov 2009) | 6 lines
Added calculated randVars:
This provides a mechanism for user calculated random variables to be stored at the end of a trial and extracted using
the normal tools (i.e. getTaskParameters)
Added named segments:
This provides a universal mechanisms to refer to task segments by name in code, reducing confusion and allowing
easier task reformulation (i.e. adding a segment)
................
r649 | dep | 2009-11-05 14:03:33 -0500 (Thu, 05 Nov 2009) | 5 lines
moved eyedata save into the saveStimData to capture the user choice result
eat keys after calibration
added shell for drift correct
merged in additions for testEyelink
................
r652 | dep | 2009-11-18 13:19:18 -0500 (Wed, 18 Nov 2009) | 2 lines
end tracking gracefully when we don't save the data
................
r653 | dep | 2009-11-18 13:20:10 -0500 (Wed, 18 Nov 2009) | 2 lines
added eyetracker field
................
r656 | dep | 2009-12-03 15:54:30 -0500 (Thu, 03 Dec 2009) | 2 lines
commented out obsolete blocks for randstate
................
r658 | dep | 2009-12-03 17:38:53 -0500 (Thu, 03 Dec 2009) | 1 line
removed the commented task rand state protection lines
................
r672 | dep | 2009-12-30 12:38:35 -0500 (Wed, 30 Dec 2009) | 2 lines
clear old screenParams after parsing
................
r674 | dep | 2009-12-30 16:34:04 -0500 (Wed, 30 Dec 2009) | 4 lines
merged in mglMake from trunk (digio)
removed 10.6 dependency by allowing __carbon__ to build without CG 10.6 full
screen call
................
r682 | dep | 2010-02-02 17:12:49 -0500 (Tue, 02 Feb 2010) | 1 line
Moved start recording from task to block callback.
................
r683 | dep | 2010-02-06 18:24:35 -0500 (Sat, 06 Feb 2010) | 2 lines
new calibration file for the g5 stim comp
................
r686 | dep | 2010-02-19 10:07:37 -0500 (Fri, 19 Feb 2010) | 3 lines
allow compile on 10.4
mgltestcal initScreens and load a low contrast grey
................
r692 | dep | 2010-02-23 14:51:54 -0500 (Tue, 23 Feb 2010) | 1 line
Updated to handle (hopefully) multiple tasks and remove the 'reset' at block boundaries until it can be implemented in a more coherent way.
................
r693 | dep | 2010-02-23 16:02:16 -0500 (Tue, 23 Feb 2010) | 2 lines
Allows a task to specify whether to stop recording at block, task or trial boundaries; allows for multiple tasks so long as only one task (optionally with multiple phases) has the 'collectEyeData' field set.
BY DEFAULT, EVEN IF YOU INIT THE EYE TRACKER, NO DATA IS RECORDED...
................
r700 | dep | 2010-02-24 11:27:13 -0500 (Wed, 24 Feb 2010) | 1 line
fixed typo
................
383 lines of code changed in 11 files: