Devakumar, Since you got the mris_make_surfaces binary as part of the 5.3.0-HCP version of FreeSurfer, it is "dynamically linked to shared libraries", which means that it is looking for libraries on your system that it needs to perform some of its work (instead of having all the code from those libraries included in the mris_make_surfaces binary itself.) The algorithm that Linux uses at runtime to locate the necessary libraries involves looking in some standard directory locations. But one can get Linux to look elsewhere before looking in the standard locations by including a list of directory names in an environment variable named LD_LIBRARY_PATH.
Through some sequence of library dependencies, mris_make_surfaces needs to be able to find a library called libnetcdf.so, and apparently, it is in this libnetcdf.so library that the symbol ncerr is defined. Since your run of mris_make_surfaces is not complaining about not being able to find libnetcdf.so, it is probably finding "a" version of that library. But apparently it is not finding the "right" version of that library. First, we should try to make sure that the right version of the libnetcdf.so library is installed on your system. Question 15 in the HCP Pipelines FAQ[1] addresses the problem of mris_make_surfaces not being able to locate the libnetcdf.so library that it needs. After reading the post linked to by the answer in that FAQ, you'll probably want to get a version of libnetcdf.so that is known to work with mris_make_surfaces. Here is a copy of the instructions I've previously given out for getting a version of the library that is known to work. We got the library from EPEL (Extra Packages for Enterprise Linux) https://fedoraproject.org/wiki/EPEL The RPM that you need which contains the shared library is netcdf-4.1.1-3.el6.5.x86_64 (assuming you are running on a 64-bit Intel platform). You should be able to get that RPM from: https://dl.fedoraproject.org/pub/epel/6/x86_64/repoview/netcdf.html[2] The thing about the RPM is that it will contain a lot more than just the libnetcdf.so file that you need, and it is intended to be installed as an entire package using the rpm tool by a system admin (someone who can login as root or sudo to root). You will need to extract the contents of the RPM, which you can find instructions for doing at http://www.cyberciti.biz/tips/how-to-extract-an-rpm-package-without-installing-it.html[3] The command you will probably end up executing based on those instructions is: $ rpm2cpio php-5.1.4-1.esp1.x86_64.rpm | cpio –idmv This should behave sort of like unpacking a tar file into a bunch of separate files. You will want to get the extracted file named libnetcdf.so.6.0.0. After consulting with your system admin about installing that library in one of the standard locations, you may find yourself in a situation in which you must "install" this library somewhere else yourself and add that somewhere else to your LD_LIBRARY_PATH. You'll probably want to create a special directory that you add as an entry to your LD_LIBRARY_PATH, put the libnetcdf.so.6.0.0 file in that directory, and finally create a symbolic link named libnetcdf.so.6 that is linked to the libnetcdf.so.6.0.0 file. It's not absolutely required, but it's pretty conventional for the actual library found when searching through a library directory (whether in a standard location or one listed in LD_LIBRARY_PATH) to be a symbolic link to the actual library used. So in that case, there really isn't a shared library file named libnetcdf.so.6. Hope that helps. Another thing to consider if the above doesn't resolve the issue: Matlab is known for setting LD_LIBRARY_PATH when it is invoked, and it may be setting LD_LIBRARY_PATH so that an old version of a library that mris_make_surfaces needs is being found. So if you've been running Matlab, it *may* be changing your LD_LIBRARY_PATH and forcing the version of libnetcdf.so that it wants to be found to be found also by mris_make_surfaces. Hope this is helpful, Tim On Thu, Jan 29, 2015, at 01:02, dkumar.in...@gmail.com wrote: > Hi > > Thank you. It is CentOS 6.5(Final). Kernel 2.6.32-431.3.1.el6.x86_64 > > Regards Devakumar > > Sent from Windows Mail > > From: Glasser, Matthew[4] Sent: Thursday, January 29, 2015 > 11:58 AM To: Devakumar gmail[5], hcp-users@humanconnectome.org > Cc:Brown, Tim[6] > > What OS is this? It’s probably too old. > > Peace, > > Matt. > > From: "dkumar.in...@gmail.com" <dkumar.in...@gmail.com> Date: > Wednesday, January 28, 2015 at 11:30 PM To: > "hcp-users@humanconnectome.org" <hcp-users@humanconnectome.org> > Subject: [HCP-Users] mris_make_surfaces: undefined symbol ncerr in > FreesurferPipelineBatch.sh > > Dear users > > I got the error message “mris_make_surfaces: symbol lookup error: > mris_make_surfaces: undefined symbol: ncerr” when running the > FreesurferPipelineBatch.sh. Can anyone help me to get out of this > problem. > > Regards Devakumar > > Sent from Windows Mail > > From: hcp-users-requ...@humanconnectome.org Sent: Monday, January 26, > 2015 11:30 PM To: hcp-users@humanconnectome.org > > Send HCP-Users mailing list submissions to > hcp-users@humanconnectome.org > > To subscribe or unsubscribe via the World Wide Web, visit > http://lists.humanconnectome.org/mailman/listinfo/hcp-users > or, via email, send a message with subject or body 'help' to > hcp-users-requ...@humanconnectome.org > > You can reach the person managing the list at > hcp-users-ow...@humanconnectome.org > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of HCP-Users digest..." > > > Today's Topics: > > 1. Re: 7 RSN Netoworks label file (Timothy Coalson) > 2. Node timeseries - How many ROI's to use? (David Hofmann) > 3. Re: Node timeseries - How many ROI's to use? (Stephen Smith) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Mon, 26 Jan 2015 01:37:20 -0600 > From: Timothy Coalson <tsc...@mst.edu> > Subject: Re: [HCP-Users] 7 RSN Netoworks label file > To: Marta Moreno <mmorenoort...@icloud.com> > Cc: "hcp-users@humanconnectome.org" <hcp-users@humanconnectome.org> > Message-ID: > <CAK_=tawv-W4swzC0yrgk7iXjEn5xvTP1RkGvFr8GPB0Y7=d...@mail.gmail.com> > Content-Type: text/plain; charset="utf-8" > > The map name/number is not a label name, it is the map index or name that > you select in the overlays toolbox in the gui, use "1" and it will probably > give you what you want. A single map usually has multiple labels, > specifically all the ones that get displayed at the same time in the gui. > > Tim > > > On Sat, Jan 24, 2015 at 10:20 PM, Marta Moreno <mmorenoort...@icloud.com> > wrote: > > > I was doing: wb_command -cifti-label-export-table but the map from > > RSN-netwoks.32k was not recognized. > > I did what Donna suggested and I got the information that way so the > > problem is resolved, but still can not find which is the map from > RSN just > > in case someone has that information. > > > > I want to extract the connectivity from my seed to different networks, > > that is why I needed the table to know which networks to extract > in matlab. > > I did it that way and is working. > > > > Thanks, > > _______________ > > Marta Moreno-Ortega, Ph.D. > > Postdoctoral Research Fellow > > Division of Experimental Therapeutics > > New York State Psychiatric Institute > > Department of Psychiatry > > Columbia University College of Physicians and Surgeons > > > > > > > > > > On Jan 23, 2015, at 12:03 AM, "Glasser, Matthew" > <glass...@wusm.wustl.edu> > > wrote: > > > > You can also use the map number if that is the issue. Else perhaps we > > need more info on what you are trying to do. > > > > Peace, > > > > Matt. > > > > On 1/22/15, 9:20 PM, "Marta Moreno" <mmorenoort...@icloud.com> wrote: > > > > Hi users, > > > > I am trying to get the labels from 7 RSN Networks but "map" on export > > table command is not recognized and I am not sure if > frontoparietal is 6 > > and DMN 7. Any idea? > > > > Thanks > > > > Sent from my iPhone > > _______________________________________________ > > HCP-Users mailing list > > HCP-Users@humanconnectome.org > > http://lists.humanconnectome.org/mailman/listinfo/hcp-users > > > > > > > > ________________________________ > > The materials in this message are private and may contain Protected > > Healthcare Information or other information of a sensitive > nature. If you > > are not the intended recipient, be advised that any unauthorized use, > > disclosure, copying or the taking of any action in reliance on the > contents > > of this information is strictly prohibited. If you have received > this email > > in error, please immediately notify the sender via telephone or > return mail. > > > > > > _______________________________________________ > > HCP-Users mailing list > > HCP-Users@humanconnectome.org > > http://lists.humanconnectome.org/mailman/listinfo/hcp-users > > > -------------- next part -------------- > An HTML attachment was scrubbed... > URL: http://lists.humanconnectome.org/pipermail/hcp-users/attachments/20150126/26e6fcab/attachment-0001.html > > ------------------------------ > > Message: 2 > Date: Mon, 26 Jan 2015 12:35:24 +0100 > From: David Hofmann <davidhofma...@gmail.com> > Subject: [HCP-Users] Node timeseries - How many ROI's to use? > To: hcp-users <hcp-users@humanconnectome.org> > Message-ID: > <caetivj0oa0orjn59pajvvaeyrrawsunjaoxomvhzwujmj9t...@mail.gmail.com> > Content-Type: text/plain; charset="utf-8" > > Hi @all, > > I downloaded the node timeseries for the HCP subjects. There are different > amounts of "ICA-ROIs", that is 25,50,100,200,300. > > I have to make a trade-off between amount of computation time, amount of > persons to include and what might be generally "prefered" or accepted by > reviewers. > > For example, I thought of either using the 25 ROI's and analyse about 50 > people or use the 100 ROI's and analyse 10 people. Both would require > approximately the same amount of computation time. > > I personally prefer to use the 100 ROI's but on the other hand 25 ROI's are > easier to handle computationally. But then I'm wondering if 25 ROI's is > "enough". > > I know there is no general answer, but I just wanted to hear some opinions > on that from you. In other words which amount of ROIs and people (25/50 or > 100/10) would you choose and why? > > Thanks in advance > > David > -------------- next part -------------- > An HTML attachment was scrubbed... > URL: http://lists.humanconnectome.org/pipermail/hcp-users/attachments/20150126/8b1167d3/attachment-0001.html > > ------------------------------ > > Message: 3 > Date: Mon, 26 Jan 2015 11:41:13 +0000 > From: Stephen Smith <st...@fmrib.ox.ac.uk> > Subject: Re: [HCP-Users] Node timeseries - How many ROI's to use? > To: David Hofmann <davidhofma...@gmail.com> > Cc: hcp-users <hcp-users@humanconnectome.org> > Message-ID: <d7654c7b-b306-4ea6-a0a4-101f587dd...@fmrib.ox.ac.uk> > Content-Type: text/plain; charset="utf-8" > > Hi - separate from your main question, in general I have a "better feeling" about netmat analysis from the 50 and 200 node datasets than the other options. Please don't ask me to defend that (or be correct about it!) - because it's just a gut feeling from having played with the datasets a lot. > > Wrt your main question - if your final analysis involves any cross-subject modelling, then the analysis is *likely* to be most limited by subject variability - and as a result probably the single most important thing to maximise is subject numbers. > > More generally, I would say you should aim to use all the data (subjects), even if you have to wait a bit longer for the results! > > Cheers, Steve. > > > > > > On 26 Jan 2015, at 11:35, David Hofmann > <davidhofma...@gmail.com> wrote: > > > > Hi @all, > > > > I downloaded the node timeseries for the HCP subjects. There are > different amounts of "ICA-ROIs", that is 25,50,100,200,300. > > > > I have to make a trade-off between amount of computation time, amount > of persons to include and what might be generally "prefered" or > accepted by reviewers. > > > > For example, I thought of either using the 25 ROI's and analyse about > 50 people or use the 100 ROI's and analyse 10 people. Both would > require approximately the same amount of computation time. > > > > I personally prefer to use the 100 ROI's but on the other hand 25 > ROI's are easier to handle computationally. But then I'm wondering if > 25 ROI's is "enough". > > > > I know there is no general answer, but I just wanted to hear some > opinions on that from you. In other words which amount of ROIs and > people (25/50 or 100/10) would you choose and why? > > > > Thanks in advance > > > > David > > > > > > > > > > > > _______________________________________________ > > HCP-Users mailing list > > HCP-Users@humanconnectome.org > > http://lists.humanconnectome.org/mailman/listinfo/hcp-users > > > > > --------------------------------------------------------------------------- > Stephen M. Smith, Professor of Biomedical Engineering > Associate Director, Oxford University FMRIB Centre > > FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK > +44 (0) 1865 222726 (fax 222717) > st...@fmrib.ox.ac.uk http://www.fmrib.ox.ac.uk/~steve > <http://www.fmrib.ox.ac.uk/~steve> > --------------------------------------------------------------------------- > > Stop the cultural?destruction?of Tibet <http://smithinks.net/> > > > > > -------------- next part -------------- > An HTML attachment was scrubbed... > URL: http://lists.humanconnectome.org/pipermail/hcp-users/attachments/20150126/422d615b/attachment-0001.html > > ------------------------------ > > _______________________________________________ > HCP-Users mailing list > HCP-Users@humanconnectome.org > http://lists.humanconnectome.org/mailman/listinfo/hcp-users > > > End of HCP-Users Digest, Vol 26, Issue 23 > ***************************************** > _______________________________________________ > HCP-Users mailing list > HCP-Users@humanconnectome.org > http://lists.humanconnectome.org/mailman/listinfo/hcp-users > > > The materials in this message are private and may contain Protected > Healthcare Information or other information of a sensitive nature. If > you are not the intended recipient, be advised that any unauthorized > use, disclosure, copying or the taking of any action in reliance on > the contents of this information is strictly prohibited. If you have > received this email in error, please immediately notify the sender via > telephone or return mail. Links: 1. https://github.com/Washington-University/Pipelines/wiki/FAQ 2. https://mail.mir.wustl.edu/OWA/redir.aspx?C=zsrmGdRETUq97cf1ApGILKVCb09xw9EI1aVQ_gFw9_3cBrnFGMIo2L1LJdMMWCOMbQhr9JwIHeQ.&URL=https%3a%2f%2fdl.fedoraproject.org%2fpub%2fepel%2f6%2fx86_64%2frepoview%2fnetcdf.html 3. https://mail.mir.wustl.edu/OWA/redir.aspx?C=zsrmGdRETUq97cf1ApGILKVCb09xw9EI1aVQ_gFw9_3cBrnFGMIo2L1LJdMMWCOMbQhr9JwIHeQ.&URL=http%3a%2f%2fwww.cyberciti.biz%2ftips%2fhow-to-extract-an-rpm-package-without-installing-it.html 4. mailto:glass...@wusm.wustl.edu 5. mailto:dkumar.in...@gmail.com 6. mailto:bro...@mir.wustl.edu -- Timothy B. Brown Business & Technology Application Analyst III Pipeline Developer (Human Connectome Project) tbbrown(at)wustl.edu ________________________________________ The material in this message is private and may contain Protected Healthcare Information (PHI). If you are not the intended recipient, be advised that any unauthorized use, disclosure, copying or the taking of any action in reliance on the contents of this information is strictly prohibited. If you have received this email in error, please immediately notify the sender via telephone or return mail. _______________________________________________ HCP-Users mailing list HCP-Users@humanconnectome.org http://lists.humanconnectome.org/mailman/listinfo/hcp-users
