Here’s are examples of the problem. yt fails immediately with any usage. As you can see, I don’t have a lot of information here! [yt-x86_64] sagittariusb 1 % yt Abort [yt-x86_64] sagittariusb 2 % yt --help Abort [yt-x86_64] sagittariusb 3 % python Python 2.7.8 (default, Sep 28 2014, 20:55:56) [GCC 4.2.1 Compatible Apple LLVM 6.0 (clang-600.0.51)] on darwin Type "help", "copyright", "credits" or "license" for more information.
import yt Abort [yt-x86_64] sagittariusb 4 %
I tried your suggestion of using the bash activate script: [yt-x86_64] sagittariusB 40 % bash yt-x86_64/bin/activate [yt-x86_64] sagittariusB 41 % yt Abort [yt-x86_64] sagittariusB 42 % FWIW, I do note that the “Abort” appears to be appearing a lot faster with the readline problem corrected. - Mordecai Begin forwarded message:
From: Nathan Goldbaum <nathan12343@gmail.com> Subject: Re: [yt-users] install failure Date: September 28, 2014 at 21:38:36 EDT To: Mordecai-Mark Mac Low <mordecai@amnh.org> Cc: <teuben@astro.umd.edu>
In your original message you said that you were using activate.csh - can you under bash with the bash activate script?
When you say "on startup", can you copy/paste a session where the error happens? Are you saying that python is failing to start or do you get the abort error when you import yt?
On Sun, Sep 28, 2014 at 6:35 PM, Mordecai-Mark Mac Low <mordecai@amnh.org> wrote: I’m taking this off the general list for the moment…
I have replaced readline with gnureadline as you suggested, and that does eliminate the error during installation (see https://dl.dropboxusercontent.com/u/31830722/yt_install-v2.log). Unfortunately, it doesn’t change the behavior on startup: Abort, with no further errors.
At Peter Teuben’s suggestion, I also ran dtruss on this, but he saw nothing obviously wrong — just a whole lot of files being opened (around 900 at his estimate).
- Mordecai
On Sep 28, 2014, at 20:45, Nathan Goldbaum <nathan12343@gmail.com> wrote:
---------- Forwarded message ---------- From: Nathan Goldbaum <nathan12343@gmail.com> Date: Sun, Sep 28, 2014 at 5:29 PM Subject: Re: [yt-users] install failure To: Discussion of the yt analysis package <yt-users@lists.spacepope.org>
So looking at the install log, it's actually the install script that's trying to build readline, right here: https://bitbucket.org/yt_analysis/yt/src/4cf5c5369df9b012759ab979330d8e356d1...
I'm actually able to reproduce this on my OSX Mavericks laptop when I do "pip install readline". It seems that the readline package on pypi hasn't been updated since 2012: https://pypi.python.org/pypi/readline.
I never noticed this issue because I use homebrew's python, and homebrew links against the gnu readline library when they build python.
It seems the gnureadline package has supplanted the old readline package: https://pypi.python.org/pypi/gnureadline
I've gone ahead and created a pull request to make the install script install gnureadline instead of readline: https://bitbucket.org/yt_analysis/yt/pull-request/1225/install-gnureadline-p...
You can probably just modify your local copy of the install script to use "pip install gnureadline" instead of "pip install readline".
Hope that fixes your issue, thanks very much for the report.
-Nathan
On Sun, Sep 28, 2014 at 5:00 PM, Mordecai-Mark Mac Low <mordecai@amnh.org> wrote: Nathan,
I’ve posted the full install log to
https://dl.dropboxusercontent.com/u/31830722/yt_install.log
It does look like ipython is trying to compile readline.
I have the regular Mac Python installation, but would have to learn anaconda or canopy before using them for an installation, which I was hoping to avoid for the moment by using the install script.
Best,
Mordecai
On Sep 28, 2014, at 19:33, yt-users-request@lists.spacepope.org wrote:
Message: 1 Date: Sun, 28 Sep 2014 13:57:47 -0700 From: Nathan Goldbaum <nathan12343@gmail.com> To: Discussion of the yt analysis package <yt-users@lists.spacepope.org> Subject: Re: [yt-users] install failure Message-ID: <CAJXewO=G9M7V0deBGtU46s68kdKPVTqsmhEBxV+yG7-K=8HcgQ@mail.gmail.com> Content-Type: text/plain; charset="utf-8"
Hi Mordecai,
We've had some reports about incompatibilities with the install script on OS X Mavericks recently, but I don't think this error has been reported before. Unfortunately getting the install script to work portably in all cases is a labor of love and sometimes errors like these crop up.
Googling for your error returns this discussion on the readline mailing list: https://lists.gnu.org/archive/html/bug-readline/2013-12/msg00002.html. Apparently the version you're compiling needs to be patched to compile cleanly on mavericks. I haven't checked whether the patch has made its way into a readline release.
I'm not sure what exactly is compiling readline (IPython maybe?), but earlier in the install script we should have installed a pure-python version of the readline library from pypi if we were unable to import it. It would help to figure out what went wrong if you could upload the full install log somewhere.
Do you have a python environment setup on your laptop already? It's very straightforward to install yt using anaconda, canopy, or a manually configured python environment, bypassing all the bootstrapping that the install script must do. Take a look at the following page in the documentation:
http://yt-project.org/doc/installing.html#installing-yt-using-anaconda http://yt-project.org/doc/installing.html#installing-yt-using-pip-or-from-so...
Hope that's helpful.
-Nathan
On Sun, Sep 28, 2014 at 12:36 PM, Mordecai-Mark Mac Low <mordecai@amnh.org> wrote:
Hi, I am having a fatal problem installing yt on my MacBook Pro running OS X v10.9.4. I am using the install script, with the options set as follows:
# Here's where you put the HDF5 path if you like; otherwise it'll download it # and install it on its own #HDF5_DIR=/opt/local/lib
# If you need to supply arguments to the NumPy or SciPy build, supply them here # This one turns on gfortran manually: #NUMPY_ARGS="--fcompiler=gnu95" # If you absolutely can't get the fortran to work, try this: #NUMPY_ARGS="--fcompiler=fake"
INST_HG=0 # Install Mercurial or not? If hg is not already # installed, yt cannot be installed. INST_ZLIB=1 # On some systems (Kraken) matplotlib has issues with # the system zlib, which is compiled statically. # If need be, you can turn this off. INST_BZLIB=1 # On some systems, libbzip2 is missing. This can # lead to broken mercurial installations. INST_PNG=1 # Install a local libpng? Same things apply as with zlib. INST_FTYPE=1 # Install FreeType2 locally? INST_ENZO=0 # Clone a copy of Enzo? INST_SQLITE3=1 # Install a local version of SQLite3? INST_PYX=1 # Install PyX? Sometimes PyX can be problematic without a # working TeX installation. INST_0MQ=1 # Install 0mq (for IPython) and affiliated bindings? INST_ROCKSTAR=1 # Install the Rockstar halo finder? INST_SCIPY=0 # Install scipy?
After downloading and activating Xcode command line tools, I got the script to complete successfully. I activated it using the activate.csh script.
However, when I then try to start yt, it pauses for 30 seconds or so, and then issues the response ?Abort?, with no further errors. I get the same response if I try to import yt into the python (running in the yt environment). Abort, and return to the command line.
I have tried erasing the yt-x86_64 directory and reinstalling from scratch, with no change in behavior. I also examined the yt_install.log. I found a number of, presumably irrelevant, errors in the configuration tests, but also an error at the end of the long section on building Ipython (almost at the end of the script, just before readline assembly that looks possibly relevant:
gcc -dynamic -arch i386 -arch x86_64 -arch_only `/usr/bin/arch` -install_name /usr/local/lib/libreadline.6.2.dylib -current_version 6.2 -compatibility_version 6 -v -o libreadline.6.2.dylib readline.so vi_mode.so funmap.so keymaps.so parens.so search.so rltty.so complete.so bind.so isearch.so display.so signals.so util.so kill.so undo.so macro.so input.so callback.so terminal.so text.so nls.so misc.so xmalloc.so xfree.so history.so histexpand.so histfile.so histsearch.so shell.so mbutil.so tilde.so compat.so -lncurses Apple LLVM version 6.0 (clang-600.0.51) (based on LLVM 3.5svn) Target: x86_64-apple-darwin13.3.0 Thread model: posix clang: error: invalid argument '-compatibility_version 6' only allowed with '-dynamiclib' clang: error: invalid argument '-compatibility_version 6' only allowed with '-dynamiclib' make[1]: *** [libreadline.6.2.dylib] Error 1 make: [shared] Error 2 (ignored)
Any insight on how to proceed?
Best,
Mordecai
-- Mordecai-Mark Mac Low Curator & Professor +1-212-496-3443 Department of Astrophysics +1-212-769-5007 (fax) American Museum of Natural History mmaclow (Skype, Hangout) 79th St at CPW, NY, NY, 10024-5192, USA
_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
-------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.spacepope.org/pipermail/yt-users-spacepope.org/attachments/20140928/d7562d47/attachment.htm>
------------------------------
Message: 2 Date: Sun, 28 Sep 2014 19:33:51 -0400 From: Professor Maller <amaller@citytech.cuny.edu> To: yt-users@lists.spacepope.org Subject: Re: [yt-users] install failure Message-ID: <CACCSjdNPC0QVrxDooBEQ2W7Mp6KaBJxNJZ9eLvH6TG8WyD7m0Q@mail.gmail.com> Content-Type: text/plain; charset="utf-8"
Hi Mordecai,
Depending on what you want to do with yt, if you need the latest build or not, but I've had success on Mac with the anaconda distribution of python which includes yt-3.0. https://store.continuum.io/cshop/anaconda/.
Ari
On Sun, Sep 28, 2014 at 4:30 PM, <yt-users-request@lists.spacepope.org> wrote:
Send yt-users mailing list submissions to yt-users@lists.spacepope.org
To subscribe or unsubscribe via the World Wide Web, visit http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org or, via email, send a message with subject or body 'help' to yt-users-request@lists.spacepope.org
You can reach the person managing the list at yt-users-owner@lists.spacepope.org
When replying, please edit your Subject line so it is more specific than "Re: Contents of yt-users digest..."
Today's Topics:
1. Re: ProfilePlot in YT-3.0 (Nathan Goldbaum) 2. install failure (Mordecai-Mark Mac Low) 3. Re: install failure (Peter Teuben)
----------------------------------------------------------------------
Message: 1 Date: Sun, 28 Sep 2014 12:14:04 -0700 From: Nathan Goldbaum <nathan12343@gmail.com> To: Discussion of the yt analysis package <yt-users@lists.spacepope.org> Subject: Re: [yt-users] ProfilePlot in YT-3.0 Message-ID: < CAJXewOkKaMPBE02RH_VY4rNTXgOobeJ02qOOZY5PRk332GGveg@mail.gmail.com> Content-Type: text/plain; charset="utf-8"
On Sunday, September 28, 2014, John Regan <johnanthonyregan@gmail.com> wrote:
Hi Nathan,
Thanks for that. I'll have a look at what helper functions would be useful as I use the ProfilePlot routine and I can issue a PR then. I think it would be worth updating the docs with info like you just sent me.
Again, please feel free to open a pull request.
Cheers, John
On Sun, Sep 28, 2014 at 11:17 AM, Nathan Goldbaum <nathan12343@gmail.com <javascript:_e(%7B%7D,'cvml','nathan12343@gmail.com');>> wrote:
Hi John,
ProfilePlot is a wrapper around the Profile1D class, which calculates
and
stores the histograms. To access the underlying data you need to access the profile object via the profiles attribute of the ProfilePlot:
import yt ds = yt.load('IsolatedGalaxy/galaxy0030/galaxy0030') plot = yt.ProfilePlot(ds.all_data(), 'radius', 'density') profile = plot.profiles[0] print profile.x_field, profile.x print profile.field_data.keys(), profile['density']
There are probably a number of ways that the current API can be improved to make it more straightforward to access this data. PRs along this front would be very welcome, I think.
-Nathan
On Sun, Sep 28, 2014 at 1:05 AM, John Regan <johnanthonyregan@gmail.com <javascript:_e(%7B%7D,'cvml','johnanthonyregan@gmail.com');>> wrote:
> HI All, > > Is there an easy way to access the x_field and the y_fields in the > ProfilePlot function? > > So for example > > prof = ProfilePlot(sphere, "radius", "density") > > > print prof.x_field > print prof.y_field[0] > > I inserted a couple of hack functions into the ProfilePlot class to > return the x_field and the .items() tuple but there is probably a better > way? > > John > > _______________________________________________ > yt-users mailing list > yt-users@lists.spacepope.org > <javascript:_e(%7B%7D,'cvml','yt-users@lists.spacepope.org');> > http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org > >
_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org <javascript:_e(%7B%7D,'cvml','yt-users@lists.spacepope.org');> http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
-------------- next part -------------- An HTML attachment was scrubbed... URL: < http://lists.spacepope.org/pipermail/yt-users-spacepope.org/attachments/2014...
------------------------------
Message: 2 Date: Sun, 28 Sep 2014 15:36:55 -0400 From: Mordecai-Mark Mac Low <mordecai@amnh.org> To: <yt-users@lists.spacepope.org> Subject: [yt-users] install failure Message-ID: <C2FC80DA-9E95-476E-BB68-E5944D4597F3@amnh.org> Content-Type: text/plain; charset="windows-1252"
Hi, I am having a fatal problem installing yt on my MacBook Pro running OS X v10.9.4. I am using the install script, with the options set as follows:
# Here's where you put the HDF5 path if you like; otherwise it'll download it # and install it on its own #HDF5_DIR=/opt/local/lib
# If you need to supply arguments to the NumPy or SciPy build, supply them here # This one turns on gfortran manually: #NUMPY_ARGS="--fcompiler=gnu95" # If you absolutely can't get the fortran to work, try this: #NUMPY_ARGS="--fcompiler=fake"
INST_HG=0 # Install Mercurial or not? If hg is not already # installed, yt cannot be installed. INST_ZLIB=1 # On some systems (Kraken) matplotlib has issues with # the system zlib, which is compiled statically. # If need be, you can turn this off. INST_BZLIB=1 # On some systems, libbzip2 is missing. This can # lead to broken mercurial installations. INST_PNG=1 # Install a local libpng? Same things apply as with zlib. INST_FTYPE=1 # Install FreeType2 locally? INST_ENZO=0 # Clone a copy of Enzo? INST_SQLITE3=1 # Install a local version of SQLite3? INST_PYX=1 # Install PyX? Sometimes PyX can be problematic without a # working TeX installation. INST_0MQ=1 # Install 0mq (for IPython) and affiliated bindings? INST_ROCKSTAR=1 # Install the Rockstar halo finder? INST_SCIPY=0 # Install scipy?
After downloading and activating Xcode command line tools, I got the script to complete successfully. I activated it using the activate.csh script.
However, when I then try to start yt, it pauses for 30 seconds or so, and then issues the response ?Abort?, with no further errors. I get the same response if I try to import yt into the python (running in the yt environment). Abort, and return to the command line.
I have tried erasing the yt-x86_64 directory and reinstalling from scratch, with no change in behavior. I also examined the yt_install.log. I found a number of, presumably irrelevant, errors in the configuration tests, but also an error at the end of the long section on building Ipython (almost at the end of the script, just before readline assembly that looks possibly relevant:
gcc -dynamic -arch i386 -arch x86_64 -arch_only `/usr/bin/arch` -install_name /usr/local/lib/libreadline.6.2.dylib -current_version 6.2 -compatibility_version 6 -v -o libreadline.6.2.dylib readline.so vi_mode.so funmap.so keymaps.so parens.so search.so rltty.so complete.so bind.so isearch.so display.so signals.so util.so kill.so undo.so macro.so input.so callback.so terminal.so text.so nls.so misc.so xmalloc.so xfree.so history.so histexpand.so histfile.so histsearch.so shell.so mbutil.so tilde.so compat.so -lncurses Apple LLVM version 6.0 (clang-600.0.51) (based on LLVM 3.5svn) Target: x86_64-apple-darwin13.3.0 Thread model: posix clang: error: invalid argument '-compatibility_version 6' only allowed with '-dynamiclib' clang: error: invalid argument '-compatibility_version 6' only allowed with '-dynamiclib' make[1]: *** [libreadline.6.2.dylib] Error 1 make: [shared] Error 2 (ignored)
Any insight on how to proceed?
Best,
Mordecai
-- Mordecai-Mark Mac Low Curator & Professor +1-212-496-3443 Department of Astrophysics +1-212-769-5007 (fax) American Museum of Natural History mmaclow (Skype, Hangout) 79th St at CPW, NY, NY, 10024-5192, USA
-------------- next part -------------- An HTML attachment was scrubbed... URL: < http://lists.spacepope.org/pipermail/yt-users-spacepope.org/attachments/2014...
------------------------------
Message: 3 Date: Sun, 28 Sep 2014 16:30:01 -0400 From: Peter Teuben <teuben@astro.umd.edu> To: yt-users@lists.spacepope.org Subject: Re: [yt-users] install failure Message-ID: <54286FC9.1010003@astro.umd.edu> Content-Type: text/plain; charset="windows-1252"
just "abort" usually tells me it can't find enough memory while loading the executable. I assume you got plenty of that.
On linux I would then use 'strace' to see if the kernel tells me more. And the ldd command to see what shared libs it might need.
On a mac, there appears to be the dtruss command for strace. Me and Mac appear to be in a continues state of dislike, all I got was dtrace: failed to initialize dtrace: DTrace requires additional privileges so I'm sure somebody can tell me what that means (I don't have admin on that mac)
peter
On 09/28/2014 03:36 PM, Mordecai-Mark Mac Low wrote:
Hi, I am having a fatal problem installing yt on my MacBook Pro running OS X v10.9.4. I am using the install script, with the options set as follows:
# Here's where you put the HDF5 path if you like; otherwise it'll download it # and install it on its own #HDF5_DIR=/opt/local/lib
# If you need to supply arguments to the NumPy or SciPy build, supply them here # This one turns on gfortran manually: #NUMPY_ARGS="--fcompiler=gnu95" # If you absolutely can't get the fortran to work, try this: #NUMPY_ARGS="--fcompiler=fake"
INST_HG=0 # Install Mercurial or not? If hg is not already # installed, yt cannot be installed. INST_ZLIB=1 # On some systems (Kraken) matplotlib has issues with # the system zlib, which is compiled statically. # If need be, you can turn this off. INST_BZLIB=1 # On some systems, libbzip2 is missing. This can # lead to broken mercurial installations. INST_PNG=1 # Install a local libpng? Same things apply as with zlib. INST_FTYPE=1 # Install FreeType2 locally? INST_ENZO=0 # Clone a copy of Enzo? INST_SQLITE3=1 # Install a local version of SQLite3? INST_PYX=1 # Install PyX? Sometimes PyX can be problematic without a # working TeX installation. INST_0MQ=1 # Install 0mq (for IPython) and affiliated bindings? INST_ROCKSTAR=1 # Install the Rockstar halo finder? INST_SCIPY=0 # Install scipy?
After downloading and activating Xcode command line tools, I got the script to complete successfully. I activated it using the activate.csh script.
However, when I then try to start yt, it pauses for 30 seconds or so, and then issues the response ?Abort?, with no further errors. I get the same response if I try to import yt into the python (running in the yt environment). Abort, and return to the command line.
I have tried erasing the yt-x86_64 directory and reinstalling from scratch, with no change in behavior. I also examined the yt_install.log. I found a number of, presumably irrelevant, errors in the configuration tests, but also an error at the end of the long section on building Ipython (almost at the end of the script, just before readline assembly that looks possibly relevant:
gcc -dynamic -arch i386 -arch x86_64 -arch_only `/usr/bin/arch` -install_name /usr/local/lib/libreadline.6.2.dylib -current_version 6.2 -compatibility_version 6 -v -o libreadline.6.2.dylib readline.so vi_mode.so funmap.so keymaps.so parens.so search.so rltty.so complete.so bind.so isearch.so display.so signals.so util.so kill.so undo.so macro.so input.so callback.so terminal.so text.so nls.so misc.so xmalloc.so xfree.so history.so histexpand.so histfile.so histsearch.so shell.so mbutil.so tilde.so compat.so -lncurses Apple LLVM version 6.0 (clang-600.0.51) (based on LLVM 3.5svn) Target: x86_64-apple-darwin13.3.0 Thread model: posix clang: error: invalid argument '-compatibility_version 6' only allowed with '-dynamiclib' clang: error: invalid argument '-compatibility_version 6' only allowed with '-dynamiclib' make[1]: *** [libreadline.6.2.dylib] Error 1 make: [shared] Error 2 (ignored)
Any insight on how to proceed?
Best,
Mordecai
-- Mordecai-Mark Mac Low Curator & Professor +1-212-496-3443 Department of Astrophysics +1-212-769-5007 (fax) American Museum of Natural History mmaclow (Skype, Hangout) 79th St at CPW, NY, NY, 10024-5192, USA
_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
-------------- next part -------------- An HTML attachment was scrubbed... URL: < http://lists.spacepope.org/pipermail/yt-users-spacepope.org/attachments/2014...
------------------------------
Subject: Digest Footer
_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
------------------------------
End of yt-users Digest, Vol 79, Issue 40 ****************************************
-------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.spacepope.org/pipermail/yt-users-spacepope.org/attachments/20140928/f001e5bc/attachment.htm>
------------------------------
Subject: Digest Footer
_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
------------------------------
End of yt-users Digest, Vol 79, Issue 41 ****************************************
-- Mordecai-Mark Mac Low Curator & Professor +1-212-496-3443 Department of Astrophysics +1-212-769-5007 (fax) American Museum of Natural History mmaclow (Skype, Hangout) 79th St at CPW, NY, NY, 10024-5192, USA
_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
-- Mordecai-Mark Mac Low Curator & Professor +1-212-496-3443 Department of Astrophysics +1-212-769-5007 (fax) American Museum of Natural History mmaclow (Skype, Hangout) 79th St at CPW, NY, NY, 10024-5192, USA
-- Mordecai-Mark Mac Low Curator & Professor +1-212-496-3443 Department of Astrophysics +1-212-769-5007 (fax) American Museum of Natural History mmaclow (Skype, Hangout) 79th St at CPW, NY, NY, 10024-5192, USA
On Sun, Sep 28, 2014 at 6:57 PM, Mordecai-Mark Mac Low <mordecai@amnh.org> wrote:
Here’s are examples of the problem. yt fails immediately with any usage. As you can see, I don’t have a lot of information here!
[yt-x86_64] sagittariusb 1 % yt Abort [yt-x86_64] sagittariusb 2 % yt --help Abort [yt-x86_64] sagittariusb 3 % python Python 2.7.8 (default, Sep 28 2014, 20:55:56) [GCC 4.2.1 Compatible Apple LLVM 6.0 (clang-600.0.51)] on darwin Type "help", "copyright", "credits" or "license" for more information.
import yt Abort [yt-x86_64] sagittariusb 4 %
I've never seen an "abort" error like this, sorry I can't be more helpful in debugging the issue you're having here.
I tried your suggestion of using the bash activate script:
[yt-x86_64] sagittariusB 40 % bash yt-x86_64/bin/activate [yt-x86_64] sagittariusB 41 % yt Abort [yt-x86_64] sagittariusB 42 %
I think you're still inside a csh/tcsh prompt (which shell are you using?). For what it's worth, I don't think too many people use the csh activate script, so there might very well be an issue with that. To check whether yt works under bash, you'll need to go into a bash prompt and then source the activate script: <tcsh>$ bash <bash>$ source yt-x86_64/bin/activate <bash>$ yt
FWIW, I do note that the “Abort” appears to be appearing a lot faster with the readline problem corrected.
- Mordecai
Begin forwarded message:
*From: *Nathan Goldbaum <nathan12343@gmail.com> *Subject: **Re: [yt-users] install failure* *Date: *September 28, 2014 at 21:38:36 EDT *To: *Mordecai-Mark Mac Low <mordecai@amnh.org> *Cc: *<teuben@astro.umd.edu>
In your original message you said that you were using activate.csh - can you under bash with the bash activate script?
When you say "on startup", can you copy/paste a session where the error happens? Are you saying that python is failing to start or do you get the abort error when you import yt?
On Sun, Sep 28, 2014 at 6:35 PM, Mordecai-Mark Mac Low <mordecai@amnh.org> wrote:
I’m taking this off the general list for the moment…
I have replaced readline with gnureadline as you suggested, and that does eliminate the error during installation (see https://dl.dropboxusercontent.com/u/31830722/yt_install-v2.log). Unfortunately, it doesn’t change the behavior on startup: Abort, with no further errors.
At Peter Teuben’s suggestion, I also ran dtruss on this, but he saw nothing obviously wrong — just a whole lot of files being opened (around 900 at his estimate).
- Mordecai
On Sep 28, 2014, at 20:45, Nathan Goldbaum <nathan12343@gmail.com> wrote:
---------- Forwarded message ---------- From: Nathan Goldbaum <nathan12343@gmail.com> Date: Sun, Sep 28, 2014 at 5:29 PM Subject: Re: [yt-users] install failure To: Discussion of the yt analysis package <yt-users@lists.spacepope.org>
So looking at the install log, it's actually the install script that's trying to build readline, right here: https://bitbucket.org/yt_analysis/yt/src/4cf5c5369df9b012759ab979330d8e356d1...
I'm actually able to reproduce this on my OSX Mavericks laptop when I do "pip install readline". It seems that the readline package on pypi hasn't been updated since 2012: https://pypi.python.org/pypi/readline.
I never noticed this issue because I use homebrew's python, and homebrew links against the gnu readline library when they build python.
It seems the gnureadline package has supplanted the old readline package: https://pypi.python.org/pypi/gnureadline
I've gone ahead and created a pull request to make the install script install gnureadline instead of readline: https://bitbucket.org/yt_analysis/yt/pull-request/1225/install-gnureadline-p...
You can probably just modify your local copy of the install script to use "pip install gnureadline" instead of "pip install readline".
Hope that fixes your issue, thanks very much for the report.
-Nathan
On Sun, Sep 28, 2014 at 5:00 PM, Mordecai-Mark Mac Low <mordecai@amnh.org
wrote:
Nathan,
I’ve posted the full install log to
https://dl.dropboxusercontent.com/u/31830722/yt_install.log
It does look like ipython is trying to compile readline.
I have the regular Mac Python installation, but would have to learn anaconda or canopy before using them for an installation, which I was hoping to avoid for the moment by using the install script.
Best,
Mordecai
On Sep 28, 2014, at 19:33, yt-users-request@lists.spacepope.org wrote:
Message: 1 Date: Sun, 28 Sep 2014 13:57:47 -0700 From: Nathan Goldbaum <nathan12343@gmail.com> To: Discussion of the yt analysis package <yt-users@lists.spacepope.org> Subject: Re: [yt-users] install failure Message-ID: <CAJXewO=G9M7V0deBGtU46s68kdKPVTqsmhEBxV+yG7-K= 8HcgQ@mail.gmail.com> Content-Type: text/plain; charset="utf-8"
Hi Mordecai,
We've had some reports about incompatibilities with the install script on OS X Mavericks recently, but I don't think this error has been reported before. Unfortunately getting the install script to work portably in all cases is a labor of love and sometimes errors like these crop up.
Googling for your error returns this discussion on the readline mailing list: https://lists.gnu.org/archive/html/bug-readline/2013-12/msg00002.html. Apparently the version you're compiling needs to be patched to compile cleanly on mavericks. I haven't checked whether the patch has made its way into a readline release.
I'm not sure what exactly is compiling readline (IPython maybe?), but earlier in the install script we should have installed a pure-python version of the readline library from pypi if we were unable to import it. It would help to figure out what went wrong if you could upload the full install log somewhere.
Do you have a python environment setup on your laptop already? It's very straightforward to install yt using anaconda, canopy, or a manually configured python environment, bypassing all the bootstrapping that the install script must do. Take a look at the following page in the documentation:
http://yt-project.org/doc/installing.html#installing-yt-using-anaconda
http://yt-project.org/doc/installing.html#installing-yt-using-pip-or-from-so...
Hope that's helpful.
-Nathan
On Sun, Sep 28, 2014 at 12:36 PM, Mordecai-Mark Mac Low <
wrote:
Hi, I am having a fatal problem installing yt on my MacBook Pro running OS X v10.9.4. I am using the install script, with the options set as follows:
# Here's where you put the HDF5 path if you like; otherwise it'll download it # and install it on its own #HDF5_DIR=/opt/local/lib
# If you need to supply arguments to the NumPy or SciPy build, supply
here # This one turns on gfortran manually: #NUMPY_ARGS="--fcompiler=gnu95" # If you absolutely can't get the fortran to work, try this: #NUMPY_ARGS="--fcompiler=fake"
INST_HG=0 # Install Mercurial or not? If hg is not already # installed, yt cannot be installed. INST_ZLIB=1 # On some systems (Kraken) matplotlib has issues with # the system zlib, which is compiled statically. # If need be, you can turn this off. INST_BZLIB=1 # On some systems, libbzip2 is missing. This can # lead to broken mercurial installations. INST_PNG=1 # Install a local libpng? Same things apply as with zlib. INST_FTYPE=1 # Install FreeType2 locally? INST_ENZO=0 # Clone a copy of Enzo? INST_SQLITE3=1 # Install a local version of SQLite3? INST_PYX=1 # Install PyX? Sometimes PyX can be problematic without a # working TeX installation. INST_0MQ=1 # Install 0mq (for IPython) and affiliated bindings? INST_ROCKSTAR=1 # Install the Rockstar halo finder? INST_SCIPY=0 # Install scipy?
After downloading and activating Xcode command line tools, I got the script to complete successfully. I activated it using the activate.csh script.
However, when I then try to start yt, it pauses for 30 seconds or so, and then issues the response ?Abort?, with no further errors. I get the same response if I try to import yt into the python (running in the yt environment). Abort, and return to the command line.
I have tried erasing the yt-x86_64 directory and reinstalling from scratch, with no change in behavior. I also examined the yt_install.log. I found a number of, presumably irrelevant, errors in the configuration tests, but also an error at the end of the long section on building Ipython (almost at the end of the script, just before readline assembly that looks possibly relevant:
gcc -dynamic -arch i386 -arch x86_64 -arch_only `/usr/bin/arch` -install_name /usr/local/lib/libreadline.6.2.dylib -current_version 6.2 -compatibility_version 6 -v -o libreadline.6.2.dylib readline.so vi_mode.so funmap.so keymaps.so parens.so search.so rltty.so complete.so bind.so isearch.so display.so signals.so util.so kill.so undo.so macro.so input.so callback.so terminal.so text.so nls.so misc.so xmalloc.so xfree.so history.so histexpand.so histfile.so histsearch.so shell.so mbutil.so tilde.so compat.so -lncurses Apple LLVM version 6.0 (clang-600.0.51) (based on LLVM 3.5svn) Target: x86_64-apple-darwin13.3.0 Thread model: posix clang: error: invalid argument '-compatibility_version 6' only allowed with '-dynamiclib' clang: error: invalid argument '-compatibility_version 6' only allowed with '-dynamiclib' make[1]: *** [libreadline.6.2.dylib] Error 1 make: [shared] Error 2 (ignored)
Any insight on how to proceed?
Best,
Mordecai
-- Mordecai-Mark Mac Low Curator & Professor +1-212-496-3443 Department of Astrophysics +1-212-769-5007 (fax) American Museum of Natural History mmaclow (Skype, Hangout) 79th St at CPW, NY, NY, 10024-5192, USA
_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
-------------- next part -------------- An HTML attachment was scrubbed... URL: < http://lists.spacepope.org/pipermail/yt-users-spacepope.org/attachments/2014...
------------------------------
Message: 2 Date: Sun, 28 Sep 2014 19:33:51 -0400 From: Professor Maller <amaller@citytech.cuny.edu> To: yt-users@lists.spacepope.org Subject: Re: [yt-users] install failure Message-ID: < CACCSjdNPC0QVrxDooBEQ2W7Mp6KaBJxNJZ9eLvH6TG8WyD7m0Q@mail.gmail.com> Content-Type: text/plain; charset="utf-8"
Hi Mordecai,
Depending on what you want to do with yt, if you need the latest build or not, but I've had success on Mac with the anaconda distribution of
which includes yt-3.0. https://store.continuum.io/cshop/anaconda/.
Ari
On Sun, Sep 28, 2014 at 4:30 PM, <yt-users-request@lists.spacepope.org
wrote:
Send yt-users mailing list submissions to yt-users@lists.spacepope.org
To subscribe or unsubscribe via the World Wide Web, visit http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org or, via email, send a message with subject or body 'help' to yt-users-request@lists.spacepope.org
You can reach the person managing the list at yt-users-owner@lists.spacepope.org
When replying, please edit your Subject line so it is more specific than "Re: Contents of yt-users digest..."
Today's Topics:
1. Re: ProfilePlot in YT-3.0 (Nathan Goldbaum) 2. install failure (Mordecai-Mark Mac Low) 3. Re: install failure (Peter Teuben)
----------------------------------------------------------------------
Message: 1 Date: Sun, 28 Sep 2014 12:14:04 -0700 From: Nathan Goldbaum <nathan12343@gmail.com> To: Discussion of the yt analysis package <yt-users@lists.spacepope.org> Subject: Re: [yt-users] ProfilePlot in YT-3.0 Message-ID: < CAJXewOkKaMPBE02RH_VY4rNTXgOobeJ02qOOZY5PRk332GGveg@mail.gmail.com> Content-Type: text/plain; charset="utf-8"
On Sunday, September 28, 2014, John Regan <johnanthonyregan@gmail.com
wrote:
Hi Nathan,
Thanks for that. I'll have a look at what helper functions would be useful as I use the ProfilePlot routine and I can issue a PR then. I think it would be worth updating the docs with info like you just sent me.
Again, please feel free to open a pull request.
Cheers, John
On Sun, Sep 28, 2014 at 11:17 AM, Nathan Goldbaum <
nathan12343@gmail.com
<javascript:_e(%7B%7D,'cvml','nathan12343@gmail.com');>> wrote:
> Hi John, > > ProfilePlot is a wrapper around the Profile1D class, which calculates and > stores the histograms. To access the underlying data you need to access > the profile object via the profiles attribute of the ProfilePlot: > > import yt > ds = yt.load('IsolatedGalaxy/galaxy0030/galaxy0030') > plot = yt.ProfilePlot(ds.all_data(), 'radius', 'density') > profile = plot.profiles[0] > print profile.x_field, profile.x > print profile.field_data.keys(), profile['density'] > > There are probably a number of ways that the current API can be improved > to make it more straightforward to access this data. PRs along this front > would be very welcome, I think. > > -Nathan > > On Sun, Sep 28, 2014 at 1:05 AM, John Regan < johnanthonyregan@gmail.com > <javascript:_e(%7B%7D,'cvml','johnanthonyregan@gmail.com');>> wrote: > >> HI All, >> >> Is there an easy way to access the x_field and the y_fields in the >> ProfilePlot function? >> >> So for example >> >> prof = ProfilePlot(sphere, "radius", "density") >> >> >> print prof.x_field >> print prof.y_field[0] >> >> I inserted a couple of hack functions into the ProfilePlot class to >> return the x_field and the .items() tuple but there is probably a better >> way? >> >> John >> >> _______________________________________________ >> yt-users mailing list >> yt-users@lists.spacepope.org >> <javascript:_e(%7B%7D,'cvml','yt-users@lists.spacepope.org');> >> http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org >> >> > > _______________________________________________ > yt-users mailing list > yt-users@lists.spacepope.org > <javascript:_e(%7B%7D,'cvml','yt-users@lists.spacepope.org');> > http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org > >
-------------- next part -------------- An HTML attachment was scrubbed... URL: <
http://lists.spacepope.org/pipermail/yt-users-spacepope.org/attachments/2014...
------------------------------
Message: 2 Date: Sun, 28 Sep 2014 15:36:55 -0400 From: Mordecai-Mark Mac Low <mordecai@amnh.org> To: <yt-users@lists.spacepope.org> Subject: [yt-users] install failure Message-ID: <C2FC80DA-9E95-476E-BB68-E5944D4597F3@amnh.org> Content-Type: text/plain; charset="windows-1252"
Hi, I am having a fatal problem installing yt on my MacBook Pro running OS X v10.9.4. I am using the install script, with the options set as follows:
# Here's where you put the HDF5 path if you like; otherwise it'll download it # and install it on its own #HDF5_DIR=/opt/local/lib
# If you need to supply arguments to the NumPy or SciPy build, supply
here # This one turns on gfortran manually: #NUMPY_ARGS="--fcompiler=gnu95" # If you absolutely can't get the fortran to work, try this: #NUMPY_ARGS="--fcompiler=fake"
INST_HG=0 # Install Mercurial or not? If hg is not already # installed, yt cannot be installed. INST_ZLIB=1 # On some systems (Kraken) matplotlib has issues with # the system zlib, which is compiled statically. # If need be, you can turn this off. INST_BZLIB=1 # On some systems, libbzip2 is missing. This can # lead to broken mercurial installations. INST_PNG=1 # Install a local libpng? Same things apply as with zlib. INST_FTYPE=1 # Install FreeType2 locally? INST_ENZO=0 # Clone a copy of Enzo? INST_SQLITE3=1 # Install a local version of SQLite3? INST_PYX=1 # Install PyX? Sometimes PyX can be problematic without a # working TeX installation. INST_0MQ=1 # Install 0mq (for IPython) and affiliated bindings? INST_ROCKSTAR=1 # Install the Rockstar halo finder? INST_SCIPY=0 # Install scipy?
After downloading and activating Xcode command line tools, I got the script to complete successfully. I activated it using the activate.csh script.
However, when I then try to start yt, it pauses for 30 seconds or so, and then issues the response ?Abort?, with no further errors. I get the same response if I try to import yt into the python (running in the yt environment). Abort, and return to the command line.
I have tried erasing the yt-x86_64 directory and reinstalling from scratch, with no change in behavior. I also examined the yt_install.log. I found a number of, presumably irrelevant, errors in the configuration tests, but also an error at the end of the long section on building Ipython (almost at the end of the script, just before readline assembly that looks possibly relevant:
gcc -dynamic -arch i386 -arch x86_64 -arch_only `/usr/bin/arch` -install_name /usr/local/lib/libreadline.6.2.dylib -current_version 6.2 -compatibility_version 6 -v -o libreadline.6.2.dylib readline.so vi_mode.so funmap.so keymaps.so parens.so search.so rltty.so complete.so bind.so isearch.so display.so signals.so util.so kill.so undo.so macro.so input.so callback.so terminal.so text.so nls.so misc.so xmalloc.so xfree.so history.so histexpand.so histfile.so histsearch.so shell.so mbutil.so tilde.so compat.so -lncurses Apple LLVM version 6.0 (clang-600.0.51) (based on LLVM 3.5svn) Target: x86_64-apple-darwin13.3.0 Thread model: posix clang: error: invalid argument '-compatibility_version 6' only allowed with '-dynamiclib' clang: error: invalid argument '-compatibility_version 6' only allowed with '-dynamiclib' make[1]: *** [libreadline.6.2.dylib] Error 1 make: [shared] Error 2 (ignored)
Any insight on how to proceed?
Best,
Mordecai
-- Mordecai-Mark Mac Low Curator & Professor +1-212-496-3443 Department of Astrophysics +1-212-769-5007 (fax) American Museum of Natural History mmaclow (Skype, Hangout) 79th St at CPW, NY, NY, 10024-5192, USA
-------------- next part -------------- An HTML attachment was scrubbed... URL: <
http://lists.spacepope.org/pipermail/yt-users-spacepope.org/attachments/2014...
------------------------------
Message: 3 Date: Sun, 28 Sep 2014 16:30:01 -0400 From: Peter Teuben <teuben@astro.umd.edu> To: yt-users@lists.spacepope.org Subject: Re: [yt-users] install failure Message-ID: <54286FC9.1010003@astro.umd.edu> Content-Type: text/plain; charset="windows-1252"
just "abort" usually tells me it can't find enough memory while loading the executable. I assume you got plenty of that.
On linux I would then use 'strace' to see if the kernel tells me more. And the ldd command to see what shared libs it might need.
On a mac, there appears to be the dtruss command for strace. Me and Mac appear to be in a continues state of dislike, all I got was dtrace: failed to initialize dtrace: DTrace requires additional privileges so I'm sure somebody can tell me what that means (I don't have admin on that mac)
peter
On 09/28/2014 03:36 PM, Mordecai-Mark Mac Low wrote:
Hi, I am having a fatal problem installing yt on my MacBook Pro running OS X v10.9.4. I am using the install script, with the
mordecai@amnh.org> them python them options
set as follows:
# Here's where you put the HDF5 path if you like; otherwise it'll download it # and install it on its own #HDF5_DIR=/opt/local/lib
# If you need to supply arguments to the NumPy or SciPy build, supply them here # This one turns on gfortran manually: #NUMPY_ARGS="--fcompiler=gnu95" # If you absolutely can't get the fortran to work, try this: #NUMPY_ARGS="--fcompiler=fake"
INST_HG=0 # Install Mercurial or not? If hg is not already # installed, yt cannot be installed. INST_ZLIB=1 # On some systems (Kraken) matplotlib has issues with # the system zlib, which is compiled statically. # If need be, you can turn this off. INST_BZLIB=1 # On some systems, libbzip2 is missing. This can # lead to broken mercurial installations. INST_PNG=1 # Install a local libpng? Same things apply as with zlib. INST_FTYPE=1 # Install FreeType2 locally? INST_ENZO=0 # Clone a copy of Enzo? INST_SQLITE3=1 # Install a local version of SQLite3? INST_PYX=1 # Install PyX? Sometimes PyX can be problematic without a # working TeX installation. INST_0MQ=1 # Install 0mq (for IPython) and affiliated bindings? INST_ROCKSTAR=1 # Install the Rockstar halo finder? INST_SCIPY=0 # Install scipy?
After downloading and activating Xcode command line tools, I got the script to complete successfully. I activated it using the activate.csh script.
However, when I then try to start yt, it pauses for 30 seconds or so, and then issues the response ?Abort?, with no further errors. I get the same response if I try to import yt into the python (running in the yt environment). Abort, and return to the command line.
I have tried erasing the yt-x86_64 directory and reinstalling from scratch, with no change in behavior. I also examined the yt_install.log. I found a number of, presumably irrelevant, errors in the configuration tests, but also an error at the end of the long section on building Ipython (almost at the end of the script, just before readline assembly that looks possibly relevant:
gcc -dynamic -arch i386 -arch x86_64 -arch_only `/usr/bin/arch` -install_name /usr/local/lib/libreadline.6.2.dylib -current_version 6.2 -compatibility_version 6 -v -o libreadline.6.2.dylib readline.so vi_mode.so funmap.so keymaps.so parens.so search.so rltty.so complete.so bind.so isearch.so display.so signals.so util.so kill.so undo.so macro.so input.so callback.so terminal.so text.so nls.so misc.so xmalloc.so xfree.so history.so histexpand.so histfile.so histsearch.so shell.so mbutil.so tilde.so compat.so -lncurses Apple LLVM version 6.0 (clang-600.0.51) (based on LLVM 3.5svn) Target: x86_64-apple-darwin13.3.0 Thread model: posix clang: error: invalid argument '-compatibility_version 6' only allowed with '-dynamiclib' clang: error: invalid argument '-compatibility_version 6' only allowed with '-dynamiclib' make[1]: *** [libreadline.6.2.dylib] Error 1 make: [shared] Error 2 (ignored)
Any insight on how to proceed?
Best,
Mordecai
-- Mordecai-Mark Mac Low Curator & Professor +1-212-496-3443 Department of Astrophysics +1-212-769-5007 (fax) American Museum of Natural History mmaclow (Skype, Hangout) 79th St at CPW, NY, NY, 10024-5192, USA
_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
-------------- next part -------------- An HTML attachment was scrubbed... URL: <
http://lists.spacepope.org/pipermail/yt-users-spacepope.org/attachments/2014...
------------------------------
Subject: Digest Footer
_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
------------------------------
End of yt-users Digest, Vol 79, Issue 40 ****************************************
-------------- next part -------------- An HTML attachment was scrubbed... URL: < http://lists.spacepope.org/pipermail/yt-users-spacepope.org/attachments/2014...
------------------------------
Subject: Digest Footer
_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
------------------------------
End of yt-users Digest, Vol 79, Issue 41 ****************************************
-- Mordecai-Mark Mac Low Curator & Professor +1-212-496-3443 Department of Astrophysics +1-212-769-5007 (fax) American Museum of Natural History mmaclow (Skype, Hangout) 79th St at CPW, NY, NY, 10024-5192, USA
_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
-- Mordecai-Mark Mac Low Curator & Professor +1-212-496-3443 Department of Astrophysics +1-212-769-5007 (fax) American Museum of Natural History mmaclow (Skype, Hangout) 79th St at CPW, NY, NY, 10024-5192, USA
-- Mordecai-Mark Mac Low Curator & Professor +1-212-496-3443 Department of Astrophysics +1-212-769-5007 (fax) American Museum of Natural History mmaclow (Skype, Hangout) 79th St at CPW, NY, NY, 10024-5192, USA
_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
Hi Mordecai, On Sun, Sep 28, 2014 at 9:04 PM, Nathan Goldbaum <nathan12343@gmail.com> wrote:
On Sun, Sep 28, 2014 at 6:57 PM, Mordecai-Mark Mac Low <mordecai@amnh.org> wrote:
Here’s are examples of the problem. yt fails immediately with any usage. As you can see, I don’t have a lot of information here!
[yt-x86_64] sagittariusb 1 % yt Abort [yt-x86_64] sagittariusb 2 % yt --help Abort [yt-x86_64] sagittariusb 3 % python Python 2.7.8 (default, Sep 28 2014, 20:55:56) [GCC 4.2.1 Compatible Apple LLVM 6.0 (clang-600.0.51)] on darwin Type "help", "copyright", "credits" or "license" for more information.
import yt Abort [yt-x86_64] sagittariusb 4 %
My guess, which could be far off, is that it's a module yt is importing. I would try these things: import numpy import matplotlib.pyplot import h5py and see if any of those error out. If they don't, then I'll send over a script importing each component of yt which you can run which will identify which particular submodule. I suspect a symbol is unresolved or causing some issue; we often see similar behavior on systems where the linking is tricky. -Matt
participants (3)
-
Matthew Turk
-
Mordecai-Mark Mac Low
-
Nathan Goldbaum