hc-3 position/time registration

Looking at ec013.553

1) The description states that the sampling rate is 30 hertz but I can not find nor figure out how the positions register to the spike times.  Help here is much appreciated.

2) The .whl file contains 4 floats.  I assume its x1 y1 x2 y2 where one of them is the red led and the other is the green led.  In any case, what is the format?  What is red/green and what is head/body?

Many thanks to this website and to the researchers who have the vision to share neural data.

Update for others with similar issues:

Moving forward under these assumptions:

1) Initial row (including '-1' no-data rows) of .whl file corresponds to timestamp zero for the spike times.  This assumption seems reasonable since enough of the directories have .whl files with time lengths that are longer then the last spike timestamp time and often the last spike timestamp is close-ish to the last position time.

2) Green led is the head/body, while red led is the nose.  In reviewing the video files, the green led's position is more stable, while the red led leads movement.

3) Confirming how the four float values correspond to the leds will require mapping various combinations of values to position in the video file.  Have not done this step yet but will post my results.  Guessing that it is GreenX GreenY RedX RedY.

My final project is due Friday, hence using assumptions since there isn't a requirement for perfectly accurate results.  Imagine that real researchers would like confirmation of these questions, however.

Again, CRCNS is an excellent endeavor, cheers.

Previously Jeremy Dobrzanski wrote:

Looking at ec013.553

1) The description states that the sampling rate is 30 hertz but I can not find nor figure out how the positions register to the spike times.  Help here is much appreciated.

 

2) The .whl file contains 4 floats.  I assume its x1 y1 x2 y2 where one of them is the red led and the other is the green led.  In any case, what is the format?  What is red/green and what is head/body?

 

Many thanks to this website and to the researchers who have the vision to share neural data.

Previously Jeremy Dobrzanski wrote:

Update for others with similar issues:

Moving forward under these assumptions:

1) Initial row (including '-1' no-data rows) of .whl file corresponds to timestamp zero for the spike times.  This assumption seems reasonable since enough of the directories have .whl files with time lengths that are longer then the last spike timestamp time and often the last spike timestamp is close-ish to the last position time.

 

2) Green led is the head/body, while red led is the nose.  In reviewing the video files, the green led's position is more stable, while the red led leads movement.

 

3) Confirming how the four float values correspond to the leds will require mapping various combinations of values to position in the video file.  Have not done this step yet but will post my results.  Guessing that it is GreenX GreenY RedX RedY.

 

 

My final project is due Friday, hence using assumptions since there isn't a requirement for perfectly accurate results.  Imagine that real researchers would like confirmation of these questions, however.

 

Again, CRCNS is an excellent endeavor, cheers.