Can't change the number of iterations

[alexsavio]

Hi,

I am forwarding a message from a colleague who is testing many of the methods implemented here. He told me he can't change the number of iterations for any of the methods that use this parameter. He has tried many syntactical ways to set this parameter through the command line. However PETPVC always does the default number of iterations: 10.

I could check the code, but do you have any idea about this issues?

Thank you!

[bathomas]

Hi Alex,

I just built and ran the following in the test directory with debug output:

eduroam-int-dhcp-97-47-185:test bathomas$ ../bin/petpvc -i filtered.nii -m 4dmask.nii -o out10.nii.gz -p IY -x 6 -y 6 -z 6 -n 10 -d Performing iterative Yang... Start fuzziness calculation sum in region 1 = 109865 sum in region 2 = 1315 sum in region 3 = 4410 sum in region 4 = 4410 matrix: 1 0 0 0 0 1 0 0 0 0 1 0 0 0 0 1

Iteration: 1 0.0322148 1.7316 2.69035 3.58713 2 0.0027501 1.99314 3.01394 4.02053 3 0.000316362 1.99993 3.01935 4.02594 4 4.23082e-05 2.00042 3.01974 4.02634 5 6.04007e-06 2.00047 3.01979 4.02639 6 8.86906e-07 2.00048 3.01979 4.02639 7 1.31718e-07 2.00048 3.01979 4.02639 8 1.96247e-08 2.00048 3.01979 4.02639 9 2.92555e-09 2.00048 3.01979 4.02639 10 4.36165e-10 2.00048 3.01979 4.02639

eduroam-int-dhcp-97-47-185:test bathomas$ ../bin/petpvc -i filtered.nii -m 4dmask.nii -o out10.nii.gz -p IY -x 6 -y 6 -z 6 -n 5 -d Performing iterative Yang... Start fuzziness calculation sum in region 1 = 109865 sum in region 2 = 1315 sum in region 3 = 4410 sum in region 4 = 4410 matrix: 1 0 0 0 0 1 0 0 0 0 1 0 0 0 0 1

Iteration: 1 0.0322148 1.7316 2.69035 3.58713 2 0.0027501 1.99314 3.01394 4.02053 3 0.000316362 1.99993 3.01935 4.02594 4 4.23082e-05 2.00042 3.01974 4.02634 5 6.04007e-06 2.00047 3.01979 4.02639

(This is a noiseless phantom image, so you won’t see much change in the numbers)

Could it be the –k switch that is causing confusion? For ‘petpvc’ the –n switch controls iterations. For ‘pvc_’ the –i switch controls iterations.

-k is only used for the approaches that also have deconvolutions in them.

If this isn’t the problem, could you reply with some debug output?

Kind regards,

Ben

From: "Alexandre M. S." notifications@github.com Reply-To: UCL/PETPVC reply@reply.github.com Date: Wednesday, 25 January 2017 at 12:51 To: UCL/PETPVC PETPVC@noreply.github.com Cc: Subscribed subscribed@noreply.github.com Subject: [UCL/PETPVC] Can't change the number of iterations (#9)

Hi,

I am forwarding a message from a colleague who is testing many of the methods implemented here. He told me he can't change the number of iterations for any of the methods that use this parameter. He has tried many syntactical ways to set this parameter through the command line. However PETPVC always does the default number of iterations: 10.

I could check the code, but do you have any idea about this issues?

Thank you!

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHubhttps://github.com/UCL/PETPVC/issues/9, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AGX4aaXpyz8ulNOBL17Xf0dTlo_Ny5GZks5rV0VCgaJpZM4Ltcr6.

[alexsavio]

Thanks @bathomas, I will check it.

[bathomas]

Hi @alexsavio, Any update on this?

[alexsavio]

Hi @bathomas. Sorry, yes! Everything is sorted out. My colleague was using a non-iterative method, expecting to change the number of iterations. (Don't ask :) ). Cheers!