Print

Print


Hi Matthew and MArk,

I am just confirming that this also fixes the error on OSX.
Have a nice weekend!
Cheers,
Andreas

Von: Matthew Webster <[log in to unmask]>
Antworten an: FSL - FMRIB's Software Library <[log in to unmask]>
Datum: Freitag, 11. Januar 2013 17:55
An: <[log in to unmask]>
Betreff: Re: [FSL] BBR error due to wrong locale settings (SOLVED)

Dear Pavel and Andreas ,
                                                 We now make greater usage of floating-point calculations in our scripts, which unfortunately causes problems with some non-english shell localisation settings. For now we suggest setting the numeric-localisation setting to english in your shell profile when using FSL tools:

export LC_NUMERIC="en_GB.UTF-8"

and we will be forcing this locale setting in fsl.sh from the next patch onwards.

I hope this helps

Kind Regards

Matthew



Dear Mark,

I am sorry it took me so long to answer. I tried Bet and everything went fine. However, the problem with BBR is not (in my opinion) caused by float point number, since the input EPI TE in the Feat GUI is integer and the value is obviously converted to float internally when the feat script starts (i.e. when I set 41 in the log appears 0,041). I tried the BBR again today (with all updates from NeuroDebian), and I got the same error as before. It works, however, when started after the LANG variable is changed to en_GB.

Kind regards

Pavel

On 23 November 2012 16:30, Mark Jenkinson <[log in to unmask]> wrote:
No - sorry - I haven't been able to replicate this here.
I have heard about another instance on a German computer though.
Do you ever see this behaviour in other GUIs?
For example, if you try the BET GUI and change the fractional intensity threshold to some other value (e.g. 0.3), does this also cause a similar problem?

All the best,
Mark



On 23 Nov 2012, at 14:02, Pavel Hok <[log in to unmask]> wrote:

Dear Mark,

did you have any luck reproducing the error? I am still getting the same error with BBR, when I forget to change the locale (I had to add the export command to my scripts to avoid the issue).

Kind regards

Pavel

On 19 October 2012 16:51, Pavel Hok <[log in to unmask]> wrote:
Dear Mark,

it is cs_CZ.UTF-8

Thanks for looking into that issue!

Pavel


On 19 October 2012 09:26, Mark Jenkinson <[log in to unmask]> wrote:
Hi,

This is very odd because we simply use calls that we've used for years without problems before.
What is your LANG variable usually set to?
I'd like to try and replicate the behaviour here.

All the best,
Mark


On 19 Oct 2012, at 08:24, Pavel Hok <[log in to unmask]> wrote:

I would like to note as well, that otherwise Feat copes well with the Czech locale settings. It is really an isolated issue of sigloss or the routine which processes the input from GUI.

Kind regards

Pavel

On 19 October 2012 09:13, Pavel Hok <[log in to unmask]> wrote:
Dear Mark,

this issue was, however, not due to wrong input, but rather an internal one. The TE is set in ms in the gui, so user cannot affect that behaviour unfortunately. The value with comma is converted internally. Only switching to English locale helps.

Kind regards

Pavel

On 12 October 2012 18:59, Mark Jenkinson <[log in to unmask]> wrote:
Dear Pavel,

This behaviour is not restricted sigloss, BBR or Feat.  
Every FSL program requires decimal points rather than commas to be used.

All the best,
Mark


On 12 Oct 2012, at 17:33, Pavel Hok <[log in to unmask]> wrote:

Dear FSL experts,

I have encountered an error after running a single subject analysis in Feat using BBR registration and B0 unwarping. I noticed this line in the main Feat script log: --te: Couldn't set_value! valstr="0,041"! It seems, that the program sigloss cannot cope with the Czech locale, where coma is used instead of decimal point. After setting export LANG="en_GB" before running feat in the same shell. Afterwards everything ran fine.

Feat main script

/bin/cp /tmp/feat_z1wN0E.fsf design.fsf

/usr/share/fsl/5.0/bin/feat_model design

mkdir .files;cp /usr/share/fsl/5.0/doc/fsl.css .files;ln -s /usr/share/fsl/5.0/doc/images .files/images

/usr/share/fsl/5.0/bin/fsl_sub -T 10 -l logs -N feat0_init   /usr/share/fsl/5.0/bin/feat /md4/TM-PMP/TM-PMP-888/Results/RHand1_1_fieldmap.feat/design.fsf -D /md4/TM-PMP/TM-PMP-888/Results/RHand1_1_fieldmap.feat -I 1 -init
21239

/usr/share/fsl/5.0/bin/fsl_sub -T 60 -l logs -N feat1b_reg -j 21239  /usr/share/fsl/5.0/bin/feat /md4/TM-PMP/TM-PMP-888/Results/RHand1_1_fieldmap.feat/design.fsf -D /md4/TM-PMP/TM-PMP-888/Results/RHand1_1_fieldmap.feat -I 1 -reg
21321

/usr/share/fsl/5.0/bin/fsl_sub -T 48 -l logs -N feat2_pre -j 21321,21239  /usr/share/fsl/5.0/bin/feat /md4/TM-PMP/TM-PMP-888/Results/RHand1_1_fieldmap.feat/design.fsf -D /md4/TM-PMP/TM-PMP-888/Results/RHand1_1_fieldmap.feat -I 1 -prestats
24156

/usr/share/fsl/5.0/bin/fsl_sub -T 5 -l logs -N feat3_film -j 24156  /usr/share/fsl/5.0/bin/feat /md4/TM-PMP/TM-PMP-888/Results/RHand1_1_fieldmap.feat/design.fsf -D /md4/TM-PMP/TM-PMP-888/Results/RHand1_1_fieldmap.feat -I 1 -film
25595

/usr/share/fsl/5.0/bin/fsl_sub -T 0 -l logs -N feat4_post -j 25595  /usr/share/fsl/5.0/bin/feat /md4/TM-PMP/TM-PMP-888/Results/RHand1_1_fieldmap.feat/design.fsf -D /md4/TM-PMP/TM-PMP-888/Results/RHand1_1_fieldmap.feat -poststats 0 0
25764

/usr/share/fsl/5.0/bin/fsl_sub -T 119 -l logs -N feat4_post -j 25764,21321  /usr/share/fsl/5.0/bin/feat /md4/TM-PMP/TM-PMP-888/Results/RHand1_1_fieldmap.feat/design.fsf -D /md4/TM-PMP/TM-PMP-888/Results/RHand1_1_fieldmap.feat -poststats 1 0
26422

/usr/share/fsl/5.0/bin/fsl_sub -T 1 -l logs -N feat5_stop -j 24156,25595,26422,21321  /usr/share/fsl/5.0/bin/feat /md4/TM-PMP/TM-PMP-888/Results/RHand1_1_fieldmap.feat/design.fsf -D /md4/TM-PMP/TM-PMP-888/Results/RHand1_1_fieldmap.feat -stop

/usr/share/fsl/5.0/bin/mainfeatreg -d /md4/TM-PMP/TM-PMP-888/Results/RHand1_1_fieldmap.feat -l /md4/TM-PMP/TM-PMP-888/Results/RHand1_1_fieldmap.feat/logs/feat5_reg -R /md4/TM-PMP/TM-PMP-888/Results/RHand1_1_fieldmap.feat/report_unwarp.html -r /md4/TM-PMP/TM-PMP-888/Results/RHand1_1_fieldmap.feat/report_reg.html  -i /md4/TM-PMP/TM-PMP-888/Results/RHand1_1_fieldmap.feat/example_func.nii.gz  -n 10 -h /md4/TM-PMP/TM-PMP-888/Results/Mprage_brain -w  BBR -x 90 -s /usr/share/fsl/5.0/data/standard/MNI152_T1_2mm_brain -y 12 -z 90 -a /md4/TM-PMP/TM-PMP-888/Results/Fieldmap_rad -b /md4/TM-PMP/TM-PMP-888/Results/Fieldmap_brain -e 41 -f 10 -g 0.69 -p y- -w BBR
Option -d ( output directory ) selected with  argument "/md4/TM-PMP/TM-PMP-888/Results/RHand1_1_fieldmap.feat"
Option -l ( logfile )input with argument "/md4/TM-PMP/TM-PMP-888/Results/RHand1_1_fieldmap.feat/logs/feat5_reg"
Option -R ( html unwarping report ) selected with  argument "/md4/TM-PMP/TM-PMP-888/Results/RHand1_1_fieldmap.feat/report_unwarp.html"
Option -r ( html registration report ) selected with  argument "/md4/TM-PMP/TM-PMP-888/Results/RHand1_1_fieldmap.feat/report_reg.html"
Option -i ( main input ) input with argument "/md4/TM-PMP/TM-PMP-888/Results/RHand1_1_fieldmap.feat/example_func.nii.gz"
Option -n ( use nonlinear reg ) input with argument "10"
Option -h ( high-res structural image ) selected with  argument "/md4/TM-PMP/TM-PMP-888/Results/Mprage_brain"
Option -w ( highres dof ) selected with  argument "BBR"
Option -x ( highres search ) selected with  argument "90"
Option -s ( standard image ) selected with  argument "/usr/share/fsl/5.0/data/standard/MNI152_T1_2mm_brain"
Option -y ( standard dof ) selected with  argument "12"
Option -z ( standard search ) selected with  argument "90"
Option -a ( fieldmap image ) selected with  argument "/md4/TM-PMP/TM-PMP-888/Results/Fieldmap_rad"
Option -b ( field ) selected with  argument "/md4/TM-PMP/TM-PMP-888/Results/Fieldmap_brain"
Option -e ( echo time ) selected with  argument "41"
Option -f ( signal loss Threshold ) selected with  argument "10"
Option -g ( echo spacing ) selected with  argument "0.69"
Option -p ( phase encoding direction ) selected with  argument "y-"
Option -w ( highres dof ) selected with  argument "BBR"

Part of FSL (build 501)
sigloss
Copyright(c) 2004, University of Oxford (Mark Jenkinson)
Estimates signal loss from a field map (in rad/s) - possibly saved by fugue


Usage:
sigloss [options] -i b0map -m mask -s sigloss_estimate

Compulsory arguments (You MUST set one or more of):
    -i,--in    input b0 map image filename (in rad/s)
    -s,--sigloss    output signal loss image filename

Optional arguments (You may optionally specify one or more of):
    -m,--mask    input mask filename
    --te    echo time (in seconds)
    -d,--slicedir    slice direction (either x , y or z)
    -v,--verbose    switch on diagnostic messages
    -h,--help    display this message



--te: Couldn't set_value! valstr="0,041"!
Cannot open volume FM_UD_fmap_sigloss for reading!
Cannot open volume FM_UD_fmap_sigloss for reading!
Image Exception : #22 :: ERROR: Could not open image FM_UD_fmap_sigloss
terminate called after throwing an instance of 'RBD_COMMON::BaseException'
Aborted
Image Exception : #22 :: ERROR: Could not open image FM_UD_sigloss+mag
terminate called after throwing an instance of 'RBD_COMMON::BaseException'
Aborted

 Cannot open sla.png for reading
Segmentation fault
rm: nelze odstranit „sl?.png“: Adresář nebo soubor neexistuje
intial HighresImage

(standard_in) 1: syntax error
(standard_in) 1: Error: comparison in expression
(standard_in) 1: syntax error
(standard_in) 1: syntax error
(standard_in) 1: Error: comparison in expression
(standard_in) 1: syntax error
/usr/share/fsl/5.0/bin/epi_reg: 172: [: =: unexpected operator

Part of FSL (build 501)
fugue
FMRIB's Utility for Geometric Unwarping of EPIs
Copyright(c) 2000-2008, University of Oxford (Mark Jenkinson)

Usage:
fugue -i  -p  -d  -u  [options]
fugue  -i  -p  -d  -w  [options]
fugue -p  -d  --saveshift= [options]

Optional arguments (You may optionally specify one or more of):
    -i,--in    filename of input volume
    -u,--unwarp    apply unwarping and save as filename
    -w,--warp    apply forward warp and save as filename
    -p,--phasemap    filename for input phase image
    -d,--dwelltoasym    set the dwell to asym time ratio
    --dwell    set the EPI dwell time per phase-encode line - same as echo spacing - (sec)
    --asym    set the fieldmap asymmetric spin echo time (sec)
    --savefmap    filename for saving fieldmap (rad/s)
    --loadfmap    filename for loading fieldmap (rad/s)
    --saveshift    filename for saving pixel shift volume
    --loadshift    filename for reading pixel shift volume
    -m,--median    apply 2D median filtering
    --despike    apply a 2D de-spiking filter
    --nofill    do not apply gap-filling measure to the fieldmap
    --noextend    do not apply rigid-body extrapolation to the fieldmap
    --smooth2    apply 2D Gaussian smoothing of sigma N (in mm)
    -s,--smooth3    apply 3D Gaussian smoothing of sigma N (in mm)
    --poly    apply polynomial fitting of order N
    --fourier    apply Fourier (sinusoidal) fitting of order N
    --pava    apply monotonic enforcement via PAVA
    --despikethreshold    specify the threshold for de-spiking (default=3.0)
    --unwarpdir    specifies direction of warping [x/y/z/x-/y-/z-] (default y)
    --phaseconj    apply phase conjugate method of unwarping
    --icorr    apply intensity correction to unwarping (pixel shift method only)
    --icorronly    apply intensity correction only (must specify output with -u,--unwarp)
    --mask    filename for loading valid mask
    --unmaskfmap    saves the unmasked fieldmap when using --savefmap
    --unmaskshift    saves the unmasked shiftmap when using --saveshift
    --nokspace    do not use k-space forward warping
    --nocheck    turn off all checking
    -v,--verbose    switch on diagnostic messages
    -h,--help    display this message



--dwell: Couldn't set_value! valstr="0,00069"!
Image Exception : #22 :: ERROR: Could not open image example_func2highres_fieldmaprads2epi_shift
terminate called after throwing an instance of 'RBD_COMMON::BaseException'
Aborted
Image Exception : #22 :: ERROR: Could not open image example_func2highres_warp
An error occured while reading file: example_func2highres_warp
Image Exception : #22 :: ERROR: Could not open image FM_UD_fmap2epi_shift
awk: cmd. line:1: BEGIN {print  * -1.0}
awk: cmd. line:1:               ^ syntax error
Cannot open volume FM_UD_fmap2epi_shift for reading!
Image Exception : #22 :: ERROR: Could not open image FM_UD_fmap2epi_shift
terminate called after throwing an instance of 'RBD_COMMON::BaseException'
Aborted
Image Exception : #22 :: ERROR: Could not open image EF_UD_shift+mag
terminate called after throwing an instance of 'RBD_COMMON::BaseException'
Aborted

 Cannot open sla.png for reading
Segmentation fault
rm: nelze odstranit „sl?.png“: Adresář nebo soubor neexistuje
Image Exception : #22 :: ERROR: Could not open image ../example_func2highres
terminate called after throwing an instance of 'RBD_COMMON::BaseException'
Aborted

 Cannot open sla.png for reading
Segmentation fault

 Cannot open ../example_func2highres1.png for reading
Segmentation fault
/bin/rm: nelze odstranit „../example_func2highres1.png“: Adresář nebo soubor neexistuje
Image Exception : #22 :: ERROR: Could not open image ../example_func2highres
terminate called after throwing an instance of 'RBD_COMMON::BaseException'
Aborted
Image Exception : #22 :: ERROR: Could not open image ../example_func2highres
terminate called after throwing an instance of 'RBD_COMMON::BaseException'
Aborted
whirlgif Rev 3.04 (c) 1997-1999 by Hans Dinsen-Hansen
                  (c) 1995-1996 by Kevin Kadow
                  (c) 1990-1993 by Mark Podlipec
Processed 2 files.
Image Exception : #22 :: ERROR: Could not open image example_func2highres
terminate called after throwing an instance of 'RBD_COMMON::BaseException'
Aborted

 Cannot open sla.png for reading
Segmentation fault

 Cannot open example_func2highres1.png for reading
Segmentation fault
/bin/rm: nelze odstranit „example_func2highres1.png“: Adresář nebo soubor neexistuje


Kind regards

Pavel Hok
--
FMRI Lab Olomouc, Czech Republic
Faculty of Medicine and Dentistry, Palacky University Olomouc
University Hospital Olomouc
Olomouc, Czech Republic








--
Pavel Hok
-----------------------------------------
-- pavelhok.ic.cz
-- webdesign a překlady
-- tel.: 731 702 547
-- mail.: [log in to unmask]
-----------------------------------------


Pavel Hok
--
FMRI Lab Olomouc, Czech Republic
Faculty of Medicine and Dentistry, Palacky University Olomouc
University Hospital Olomouc
Olomouc, Czech Republic
fmri.upol.cz





--
Pavel Hok
-----------------------------------------
-- pavelhok.ic.cz
-- webdesign a překlady
-- tel.: 731 702 547
-- mail.: [log in to unmask]
-----------------------------------------