CRP Toolbox for Visual Fixations - Set Threshold?

General discussions and questions about recurrence plot related methods.

CRP Toolbox for Visual Fixations - Set Threshold?

Postby mdongho » Tue Aug 2, 2016 23:20

Hi,

I'm quite new to RQA and I'm trying to use the CRP Toolbox to analyze visual fixation data (eye-tracking data).
My data is organized into two columns, the first column has x-coordinates and the second column has y-coordinates, with each row representing a visual fixation in the sequence in which they occurred in the trial.
For example, for one subject I have:

448.800000000000 328.800000000000
560.100000000000 309.200000000000
506.800000000000 325.600000000000
474.600000000000 356.900000000000
491.100000000000 492.100000000000
481.200000000000 439.200000000000
478.200000000000 409.900000000000
557.600000000000 319.200000000000
490.100000000000 335
433.800000000000 343.300000000000
496.600000000000 448.900000000000
478.100000000000 506.800000000000
500.500000000000 516.600000000000
489.600000000000 454.300000000000
581.100000000000 438.300000000000
556.200000000000 318.700000000000
620.100000000000 262.200000000000
542.600000000000 278.100000000000
419.500000000000 285.900000000000
431.300000000000 307.700000000000
604.400000000000 302.600000000000
534.700000000000 362.200000000000
450.200000000000 472
499.500000000000 509.700000000000
521.600000000000 497.100000000000
501.300000000000 511.300000000000
514.700000000000 343.200000000000

I have multiple subjects and I want to create an RP for each subject, separately. I'm using the function crp2 to create a RP for my data. However, I'm having trouble with picking the threshold.

I'm hoping to have a fixed threshold value that uses a fixed Euclidean distance. Thus I used the Euclidean Norm as my Neighbourhood value, however I want this distance to be the same for each of the subjects I create an RP with. It seems like the threshold is always dependent on the standard deviation of the subject's data.

Because the standard deviation between the subjects varies, I'm hoping I could just set the threshold distance to a specific value so that the RPs can be created the same way for each trial.

Is there a method of doing this with the CRP Tool box?

Thank you!
mdongho
Junior
 
Posts: 1
Joined: Tue Aug 2, 2016 23:03

Re: CRP Toolbox for Visual Fixations - Set Threshold?

Postby Norbert » Mon Aug 29, 2016 14:47

Hi,

you have to use the 'nonorm' option in the crp-function, e.g.,
Code: Select all
r = crp2(data,m,tau,10,'nonorm');

Best wishes
Norbert
User avatar
Norbert
Expert
 
Posts: 173
Joined: Wed Jan 4, 2006 11:03
Location: Potsdam Institute for Climate Impact Research, Germany


Return to General Discussion

Who is online

Users browsing this forum: No registered users and 1 guest

cron