Hi Matthew, 

I have a short follow up question regarding Weihui's question on the design.grp file.
I read the randomise help page but am not sure if I understood the information the correct - I read it now 3 times and still am not sure how to proceed.

Say I have a Pre-Post design TBSS-Analysis with 8 subjects (I attach the design .mat and .con file - that I created - a 'normal' paired t-test design)

For a rather stupid reason subjects have been ordered like this 
01) Subj1_Post 
02) Subj2_Post
03) Subj3_Post
...
08) Subj8_Post
09) Subj1_Pre
10) Subj2_Pre
11) Subj3_Pre
...
16) Subj8_Pre

Can I use the .con and .mat file as created and how shoud the .grp file look like? Or do I need to follow a different approach?


________________________________________
Von: FSL - FMRIB's Software Library [[log in to unmask]] im Auftrag von Matthew Webster [[log in to unmask]]
Gesendet: Dienstag, 5. Januar 2010 07:48
An: [log in to unmask]
Betreff: Re: [FSL] a problem of using randomise... any help is appreciated

Hi,
     If you wish to ensure that only intra-subject permutations are allowed in your analysis then you can use an design.grp file to control the permutations ( see http://www.fmrib.ox.ac.uk/fsl/randomise/index.html for more details ).

Many Regards
Matthew

Hi,Matthew,

I have a question about comparing two groups on DTI through FDTdiffusion: mine two groups are not independent groups, they are the same subjects at different times, how can I compare their FA?They should be paired two groups. Are they the same as other ones to use randomise or someone else?


Much appreciated for your help.



weihui   li  [http://graphics.hotmail.com/emsmilep.gif]




________________________________
Date: Tue, 5 Jan 2010 15:24:14 +0000
From: [log in to unmask]<mailto:[log in to unmask]>
Subject: Re: [FSL] a problem of using randomise... any help is appreciated
To: [log in to unmask]<mailto:[log in to unmask]>

Hi,
    Randomise has been used for subject groups in excess of 1000 - if you send me your input files I should be able to
diagnose the problem with your initial analysis.

Many Regards

Matthew
I selected 10 subjects from original dataset and run the code again and it works!
Is there a maximum subject number that randomise can deal with?

--
Chunhui Chen

在2010-01-05 17:33:23，chen <[log in to unmask]<mailto:[log in to unmask]>> 写道：
Thanks!
I tried, but failed again, with the same error.


--
Chunhui Chen


在2010-01-04 22:40:13，"Matthew Webster" <[log in to unmask]<mailto:[log in to unmask]>> 写道：
>Hi,
>     Do you get the same error message ( or extra output files ) if
>you specify an additional output option ( e.g -c 3?)
>
>Many Regards
>Matthew
>
>> Dear all,
>> I am using fslvbm to compare structural images of two groups, 146
>> and 158 subjects respectively. I kept getting "unexpected errors"
>> from randominze.
>>  Bellow is the error log.  Can anyone help me fix it? Thanks a lot!
>>
>> randomise options: -i GM_mod_merg_s2 -o GM_mod_merg_s2 -m GM_mask -d
>> design.mat -t design.con -V
>> Loading Data:
>> ****************************************************************************************************************************************************************************************************************************************************************************************************************
>> Data loaded
>> 1.17684e+90 permutations required for exhaustive test of t-test 1
>> Doing 5000 random permutations
>> Warning! No output options selected. Outputing raw tstat only
>> Starting permutation 1 (Unpermuted data)
>> ERROR: Program failed, unknown exception
>> Exiting
>>
>> --
>> Chunhui Chen
>>
>>







________________________________
Your E-mail and More On-the-Go. Get Windows Live Hotmail Free. Sign up now.<http://clk.atdmt.com/GBL/go/171222985/direct/01/>