[ome-devel] Fwd: Matlab Integration with OME

[Tom Macura]

Begin forwarded message:

> From: "#HU GUANGHUI#" <S8118235C at ntu.edu.sg>
> Date: May 31, 2006 12:03:31 PM BDT
> To: "Tom Macura" <tm289 at cam.ac.uk>
> Subject: RE: Matlab Integration with OME
>
>
> Hi Tom,
>
> Thanks for pointing out the error I've made.
>
> After installing of the perl modules needed by OME, clean up my OME  
> database, everything is running smoothly (or at least that's what i  
> think). During the installation, at some point the program  
> displayed something like:
>
> Xlib: No protocol specified
>
> Xlib: connection to ":0.0" refused by server
> Xlib: No protocol specified
>
> [SUCCESS]
>
> I dont really know what it means but I'm hoping the "[SUCCESS]"  
> should indicate everything is still installed properly.
>
> So if the installation was successful, I guess the next thing is to  
> test Matlab proper. From OME's website, there's this  
> HistogramEqualizationChain.xml as well as the module that can be  
> used for testing, however, i think i'm stuck again.
>
> Questions start to appear over my head, should i first import  
> analysis chain or module or both?
> Also, from the instructions, am i right to say that in the Matlab  
> R12 and above there's a function/command called "histeq"? And in  
> order to use this function i need to first import into OME and then  
> execute this function on the datasets in OME?
>
> Based on my assumption that the flow is what described above, would  
> there be any difference when it comes to user-defined functions? Is  
> it for every function to be used, a analysis chain and module is to  
> be created? If so, is there a standard way to create them?
>
> Thank you once again!
>
> Cheers,
> GuangHui
>
> -----Original Message-----
> From: Tom Macura [mailto:tm289 at cam.ac.uk]
> Sent: Wed 5/31/2006 12:39 AM
> To: #HU GUANGHUI#
> Cc: ome-devel Development
> Subject: Re: Matlab Integration with OME
>
> Hi,
>
> You did almost everything right.
>
> A CVS checkout basically is equivalent to an OME install package. You
> have to install OME (by running perl install.pl)
> even after checkout.
>
> So you have to check out into your home directory. e.g. /home/tmacur1
> or /home/guanghui
> instead of / as you currently did.
>
> Then you have to install by running sudo perl /home/tmacur1/OME/
> install.pl
>
> so now do
>
> rm -rf /OME (To remove your current OME checkout -- it's in the wrong
> place)
> cd /home/guanghui (to move to your home directory)
> cvs -z3 -d ':pserver:anoncvs at cvs.openmicroscopy.org.uk:/ome' checkout
> OME (to do a current checkout)
> cd OME
> sudo perl install.pl  (to install OME).
>
> Let me know if you need more info or have further questions,
>
> Tom
>
>
> On May 30, 2006, at 5:13 PM, #HU GUANGHUI# wrote:
>
> >
> > Hi Tom,
> >
> > Thanks for the quick reply!
> >
> > As it is my first time using cvs, this is what i've done.
> > (I'm using fedora core 3 in case the platform matters)
> > As the root, under the "/" directory, i ran these commands:
> >
> > $ cvs -d ':pserver:anoncvs at cvs.openmicroscopy.org.uk:/ome' login
> > CVS password:anoncvs
> > $ cvs -z3 -d ':pserver:anoncvs at cvs.openmicroscopy.org.uk:/ome'
> > checkout OME
> >
> > Then the cvs did a lot of updating and checkout.
> >
> > Next I ran this, under "/OME" the directory where OME is installed
> > $ cvs update -d
> >
> > However, i saw warning messages like the following
> >
> > cvs update: Updating conf
> > cvs update: warning: cannot write history file /ome/CVSROOT/
> > history: Permission denied
> > cvs update: move away conf/httpd.ome.conf; it is in the way
> > C conf/httpd.ome.conf
> > cvs update: warning: cannot write history file /ome/CVSROOT/
> > history: Permission denied
> > cvs update: move away conf/httpd.omeds-dev.conf; it is in the way
> > C conf/httpd.ome.conf
> >
> > and the list goes on...
> > And when i did a "slocate CVSROOT", it was not able to find such
> > directory.
> > So I was afraid I did something wrong, or misunderstand the
> > instructions.
> >
> > Also, I understand OME can be "reinstalled" if so how am I going to
> > perform the reinstallation?
> > Is it by running the script "install.pl" that is found in the
> > directory of the installed OME?
> >
> > Thanks for your time, I'm also eager to get the integration up and
> > running.
> >
> > Cheers,
> > GuangHui
> >
> > -----Original Message-----
> > From: Tom Macura [mailto:tm289 at cam.ac.uk]
> > Sent: Sun 5/28/2006 3:24 AM
> > To: #HU GUANGHUI#
> > Subject: Re: Matlab Integration with OME
> >
> > Hi GuangHui,
> >
> > If you want to use OME with MATLAB you really have no choice but to
> > get the latest version of OME from CVS (info at
> > cvs.openmicroscopy.org.uk).
> > The current release (2.4.0) is over a year old. There have been many
> > many improvements to the OME/MATLAB coupling.  A new release is in
> > the pipeline but I have no idea when it will happen.
> >
> > The CVS version of OME is pretty good -- don't worry about bugs the
> > MATLAB part is very thoroughly tested (because we use it all the  
> time
> > as infrastructure for our own research) Unfortunately they are  
> always
> > unknown problems.
> >
> > You might want to look at the paper I wrote about OME/MATLAB
> > coupling:  http://www.openmicroscopy.org.uk/publications/
> > MatlabHandler.pdf
> >
> > Hope things work for you. Don't hesitate to write back. I'm very
> > excited you're interested in OME+MATLAB and want to help you,
> >
> > Best,
> >
> > Tom
> >
>
>
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.openmicroscopy.org.uk/pipermail/ome-devel/attachments/20060604/6577ec00/attachment.htm