[lsst-dm-stack-users] Welcome to the MAF tutorial mailing list
Kathy Vivas
kvivas at ctio.noao.edu
Tue Aug 5 17:19:25 PDT 2014
Lynne,
yes, the direct anaconda installation was successful, and the newinstall.sh ran with no problems at all. Yeah!!!
I am now installing the lsst_sims. This will probably take a while. I will let you know how if ends.
Thanks a lot for you help!!
Kathy
On Aug 5, 2014, at 7:35 PM, Lynne Jones <ljones.uw at gmail.com> wrote:
> Well, I have to say, I’m a bit stumped. You’re right — none of the things you’re listing there seem like they should have an effect.
>
> At this point, I’d suggest trying to install anaconda python yourself, outside the stack. Since the stack version of anaconda is a bit old, hopefully this is a problem that will go away with a more recent anaconda version.
>
> To do that, go here - https://store.continuum.io/cshop/anaconda/ - and I think you should be able to just click the big ‘download’ button.
> Once it’s finished, make sure it’s in your path as Peter mentioned (a quick ‘which python’ is a first step) and then do a quick test like starting up python and typing ‘import scipy’ would probably do the trick .. if you get an import error when you try doing ‘import scipy’, let me know and we’ll try to figure that out as it probably means something is wrong in the installation.
>
> After that, make sure you’re in a window that has anaconda in the proper path, then try the MAF installation instructions again, but when the newinstall.sh script asks if you want to install LSST-managed anaconda python, answer ‘No’.
>
> The rest of the stack should install and then MAF should run at that point but to tell the truth, I haven’t tried it myself. So, let me know if you have other problems.
>
> Lynne
>
>
>
> On Aug 5, 2014, at 4:25 PM, Kathy Vivas <kvivas at ctio.noao.edu> wrote:
>
>> ok... I have been reading the e-mails the group sent. Thanks a lot!
>>
>> To answer from the beginning:
>>
>>> - did you delete the lsst directory you were installing into between the first and second attempt? (I see the first time you ran ‘sudo bash newinstall.sh’ and the second time as ‘bash newinstall.sh’ — without sudo is the way to do it).
>>>
>>
>> yes, I deleted completely the directory and started over again
>>
>>
>>> - do you have an ARCH variable set?
>>
>> only IRAFARCH=macintel
>>
>> but I guess that does not count.
>>
>>>
>>> - how about your DYLD_LIBRARY_PATH and LD_LIBRARY_PATH?
>>>
>>
>> not exactly... only this:
>>
>> DYLD_FALLBACK_LIBRARY_PATH=/System/Library/Frameworks/ApplicationServices.framework/Versions/A/Frameworks/ImageIO.framework/Versions/A/Resources:/usr/local/scisoft///lib:/usr/local/scisoft///lib
>>
>>
>>> - is there anything present in /Users/vivas/lsst/DarwinX86/anaconda/1.8.0/pkgs/python-2.7.5-3/lib/ — specifically, is libz.1.dylib actually there? (if so, then this seems like a library path problem)
>>>
>>
>> That directory contains
>> libpython2.7.dylib pkgconfig python2.7
>>
>>
>> and no, apparently there is no file called libz.1.dylib in my system.
>>
>> BTW, yes, I migrated to Maverick just a couple months ago. This is a 2011 macbook pro.
>>
>> Thanks a lot for your time helping me to install the software
>>
>> Best,
>>
>> Kathy
>>
>>>
>>>
>>> On Aug 5, 2014, at 1:48 PM, Kathy Vivas <kvivas at ctio.noao.edu> wrote:
>>>
>>>> Hi again,
>>>>
>>>> First, it seems there are no environmental variables related with LSST and EUPS.
>>>>
>>>> So, I tried again. It still gives an error but it is a different one this time (???). The only thing different I did was to rename the directory /developer, following the advice given in the installation web page that says that if running into problems installing the software in a Mac we should try that (this was supposed to be done for problems related with phyephem, but I did it anyway).
>>>>
>>>> Attached is the log file, and this is the output in screen:
>>>>
>>>> vivas% bash newinstall.sh
>>>>
>>>> LSST Software Stack Builder
>>>> =======================================================================
>>>>
>>>> Detected git version 1.8.5.2 (Apple Git-48). OK.
>>>>
>>>>
>>>> In addition to Python 2.7, some LSST packages depend on recent versions of numpy,
>>>> matplotlib, and scipy. If you don't have all of these, the installation may fail.
>>>> Using the Anaconda Python distribution will ensure all these are set up.
>>>>
>>>> Anaconda Python installed by this installer will be managed by LSST's EUPS
>>>> package manager, and will not replace or modify your system python.
>>>>
>>>> Would you like us to install Anaconda Python distribution (if unsure, say yes)? yes
>>>>
>>>> Installing EUPS (v1.3.0)... done.
>>>> Installing Anaconda Python Distribution ...
>>>> [ 1/1 ] anaconda 1.8.0 ...
>>>>
>>>> ***** error: from /Users/vivas/lsst/EupsBuildDir/DarwinX86/anaconda-1.8.0/build.log:
>>>> installing: xlrd-0.9.2-py27_0 ...
>>>> installing: xlwt-0.7.5-py27_0 ...
>>>> installing: yaml-0.1.4-1 ...
>>>> installing: zeromq-2.2.0-1 ...
>>>> installing: zlib-1.2.7-1 ...
>>>> installing: anaconda-1.8.0-np17py27_0 ...
>>>> installing: _cache-0.0-x0 ...
>>>> Python 2.7.5 :: Continuum Analytics, Inc.
>>>> creating default environment...
>>>> Traceback (most recent call last):
>>>> File "/Users/vivas/lsst/DarwinX86/anaconda/1.8.0/pkgs/conda-2.0.4-py27_0/lib/python2.7/site-packages/conda/install.py", line 36, in <module>
>>>> import subprocess
>>>> File "/Users/vivas/lsst/DarwinX86/anaconda/1.8.0/pkgs/python-2.7.5-3/lib/python2.7/subprocess.py", line 432, in <module>
>>>> import pickle
>>>> File "/Users/vivas/lsst/DarwinX86/anaconda/1.8.0/pkgs/python-2.7.5-3/lib/python2.7/pickle.py", line 1266, in <module>
>>>> import binascii as _binascii
>>>> ImportError: dlopen(/Users/vivas/lsst/DarwinX86/anaconda/1.8.0/pkgs/python-2.7.5-3/lib/python2.7/lib-dynload/binascii.so, 2): Library not loaded: @loader_path/../../libz.1.dylib
>>>> Referenced from: /Users/vivas/lsst/DarwinX86/anaconda/1.8.0/pkgs/python-2.7.5-3/lib/python2.7/lib-dynload/binascii.so
>>>> Reason: image not found
>>>> + exit -5
>>>> eups distrib: Failed to build anaconda-1.8.0.eupspkg: Command:
>>>> source /Users/vivas/lsst/eups/bin/setups.sh; export EUPS_PATH=/Users/vivas/lsst; (/Users/vivas/lsst/EupsBuildDir/DarwinX86/anaconda-1.8.0/build.sh) >> /Users/vivas/lsst/EupsBuildDir/DarwinX86/anaconda-1.8.0/build.log 2>&1 4>/Users/vivas/lsst/EupsBuildDir/DarwinX86/anaconda-1.8.0/build.msg
>>>> exited with code 251
>>>>
>>>>
>>>>
>>>> --
>>>> This message has been scanned for viruses and
>>>> dangerous content by MailScanner, and is
>>>> believed to be clean.
>>>> <build.log>
>>>>
>>>>
>>>> Kathy
>>>>
>>>> On Aug 5, 2014, at 3:32 PM, Lynne Jones <ljones.uw at gmail.com> wrote:
>>>>
>>>>> Well, that is a most perplexing error message.
>>>>>
>>>>> I’ve just done a test installation of MAF on my mac (similar setup) and it went through okay.
>>>>>
>>>>> Would you mind trying it one more time and sending me the new build log?
>>>>> Actually - before you start installing again, could you also just send me the results of
>>>>> ‘env | grep LSST ‘
>>>>> and
>>>>> ‘env |grep EUPS’
>>>>> just to check there aren’t any pre-existing related environment variables that may be interfering? (seems unlikely unless you’ve installed part of LSST’s software stack in the past, but is something to double-check).
>>>>>
>>>>>
>>>>> Lynne
>>>>>
>>>>>
>>>>>
>>>>> On Aug 5, 2014, at 10:37 AM, Kathy Vivas <kvivas at ctio.noao.edu> wrote:
>>>>>
>>>>>> Thanks Lynne. Here it goes.
>>>>>>
>>>>>> Kathy
>>>>>>
>>>>>>
>>>>>>
>>>>>> --
>>>>>> This message has been scanned for viruses and
>>>>>> dangerous content by MailScanner, and is
>>>>>> believed to be clean.
>>>>>> <build.log>
>>>>>>
>>>>>> On Aug 5, 2014, at 1:03 PM, Lynne Jones <ljones.uw at gmail.com> wrote:
>>>>>>
>>>>>>> Hi Kathy,
>>>>>>>
>>>>>>> Can you attach the build.log file? It’s mentioned in your error message below, should be at
>>>>>>>> /Users/vivas/lsst/EupsBuildDir/DarwinX86/anaconda-1.8.0/build.log
>>>>>>>
>>>>>>> I’m not entirely sure what the problem is, if I can figure it out from the log file, I’ll let you know otherwise I’ll have to ask some other people at LSST.
>>>>>>>
>>>>>>> Lynne
>>>>>>>
>>>>>>> On Aug 5, 2014, at 8:19 AM, Kathy Vivas <kvivas at ctio.noao.edu> wrote:
>>>>>>>
>>>>>>>> Hi!
>>>>>>>>
>>>>>>>> So, I am in my first attempt of installing the MAF software in preparation for next's week workshop.
>>>>>>>> No success in my first attempt. Below is the error I got. I have a MacBook Pro running OS X 10.9.4 (Maverick). Any idea what to do? Should I try to install Anaconda separately? I am not fluent in python (shame on me!). Thanks in advance for any advice you may have.
>>>>>>>>
>>>>>>>> Cheers,
>>>>>>>>
>>>>>>>> Kathy
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> [Kathys-MacBook-Pro:~/lsst] vivas% sudo bash newinstall.sh
>>>>>>>>
>>>>>>>> LSST Software Stack Builder
>>>>>>>> =======================================================================
>>>>>>>>
>>>>>>>> Detected git version 1.8.5.2 (Apple Git-48). OK.
>>>>>>>>
>>>>>>>>
>>>>>>>> In addition to Python 2.7, some LSST packages depend on recent versions of numpy,
>>>>>>>> matplotlib, and scipy. If you don't have all of these, the installation may fail.
>>>>>>>> Using the Anaconda Python distribution will ensure all these are set up.
>>>>>>>>
>>>>>>>> Anaconda Python installed by this installer will be managed by LSST's EUPS
>>>>>>>> package manager, and will not replace or modify your system python.
>>>>>>>>
>>>>>>>> Would you like us to install Anaconda Python distribution (if unsure, say yes)? yes
>>>>>>>>
>>>>>>>> Installing EUPS (v1.3.0)... done.
>>>>>>>> Installing Anaconda Python Distribution ...
>>>>>>>> [ 1/1 ] anaconda 1.8.0 ...
>>>>>>>>
>>>>>>>> ***** error: from /Users/vivas/lsst/EupsBuildDir/DarwinX86/anaconda-1.8.0/build.log:
>>>>>>>> + tar xzvf /tmp/eups/distrib/anaconda_1r2jww
>>>>>>>> anaconda-1.8.0/
>>>>>>>> anaconda-1.8.0/ups/
>>>>>>>> anaconda-1.8.0/ups/pkginfo
>>>>>>>> anaconda-1.8.0/ups/anaconda.table
>>>>>>>> anaconda-1.8.0/ups/eupspkg.cfg.sh
>>>>>>>> anaconda-1.8.0/.gitignore
>>>>>>>> ++ find . -maxdepth 1 -type d '!' -name '.*'
>>>>>>>> ++ head -n 1
>>>>>>>> + PKGDIR=./anaconda-1.8.0
>>>>>>>> + test '!' -z ./anaconda-1.8.0
>>>>>>>> + cd ./anaconda-1.8.0
>>>>>>>> + [[ ! -e ./ups/eupspkg ]]
>>>>>>>> + mkdir -p ./ups
>>>>>>>> + ln -s /Users/vivas/lsst/eups/lib/eupspkg.sh ups/eupspkg
>>>>>>>> + eups list -s
>>>>>>>> eups list: No products found
>>>>>>>> + ./ups/eupspkg VERBOSE=0 FLAVOR=DarwinX86 fetch
>>>>>>>> + ./ups/eupspkg VERBOSE=0 FLAVOR=DarwinX86 prep
>>>>>>>> + exit -2
>>>>>>>> eups distrib: Failed to build anaconda-1.8.0.eupspkg: Command:
>>>>>>>> source /Users/vivas/lsst/eups/bin/setups.sh; export EUPS_PATH=/Users/vivas/lsst; (/Users/vivas/lsst/EupsBuildDir/DarwinX86/anaconda-1.8.0/build.sh) >> /Users/vivas/lsst/EupsBuildDir/DarwinX86/anaconda-1.8.0/build.log 2>&1 4>/Users/vivas/lsst/EupsBuildDir/DarwinX86/anaconda-1.8.0/build.msg
>>>>>>>> exited with code 254
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> On Jul 21, 2014, at 4:02 PM, Lynne Jones <ljones.uw at gmail.com> wrote:
>>>>>>>>
>>>>>>>>> Welcome to the mailing list for the Metrics Analysis Framework (MAF) tutorial session at the 2014 LSST & NOAO Observing Cadences Workshop.
>>>>>>>>>
>>>>>>>>> Just to check everyone's in the right room - MAF is the software package developed by the LSST Simulations team to help anyone and everyone quantify the performance of OpSim simulated surveys for their science goals. You’re getting this email because you expressed an interest in MAF and an intent to attend the 3-hour MAF hands-on tutorial on Monday Aug 11 at the start of the Observing Cadences Workshop. We’re looking forward to seeing you. We think we’ve got a pretty exciting session planned.
>>>>>>>>>
>>>>>>>>> Details on what the tutorial will cover are available at https://confluence.lsstcorp.org/display/SIM/MAF+Tutorials+for+Cadence+Workshop. The tutorial will be broken into units. Each unit builds on what has been learned in the previous unit, with a progression from an introduction to the system, to using a ‘stock’ version of MAF, to creating your own new analysis tools with MAF.
>>>>>>>>>
>>>>>>>>> Because this is intended to be a hands-on session and we have only a limited time together in person, we expect each of you to do some “homework” ahead of time. This homework involves downloading and installing MAF, then running two simple tests. Instructions for the homework are available at https://confluence.lsstcorp.org/display/SIM/MAF+Installation+and+Initial+Examples. We’ve even got a quiz*** for you at http://ls.st/upm
>>>>>>>>>
>>>>>>>>> Please do use this mailing list if you have any questions, problems with the installation, or other concerns, so we can all learn together.
>>>>>>>>>
>>>>>>>>> Go MAF!
>>>>>>>>> The LSST Simulations Team
>>>>>>>>>
>>>>>>>>> *** Okay, most of the ‘quiz’ is actually a survey designed to help us figure out how to best fit the tutorial to you, but there is actually a quiz question at the end. We will be pinging you to check in based on your response to that question. So, please do fill out the survey. http://ls.st/upm
>>>>>>>>>
>>>>>>>>> --
>>>>>>>>> This message has been scanned for viruses and
>>>>>>>>> dangerous content by MailScanner, and is
>>>>>>>>> believed to be clean.
>>>>>>>>
>>>>>>>>
>>>>>>>> --
>>>>>>>> This message has been scanned for viruses and
>>>>>>>> dangerous content by MailScanner, and is
>>>>>>>> believed to be clean.
>>>>>>>
>>>>>>>
>>>>>>> --
>>>>>>> This message has been scanned for viruses and
>>>>>>> dangerous content by MailScanner, and is
>>>>>>> believed to be clean.
>>>>>>
>>>>>
>>>>>
>>>>> --
>>>>> This message has been scanned for viruses and
>>>>> dangerous content by MailScanner, and is
>>>>> believed to be clean.
>>>>
>>>
>>>
>>> --
>>> This message has been scanned for viruses and
>>> dangerous content by MailScanner, and is
>>> believed to be clean.
>>
>>
>> --
>> This message has been scanned for viruses and
>> dangerous content by MailScanner, and is
>> believed to be clean.
>
>
> --
> This message has been scanned for viruses and
> dangerous content by MailScanner, and is
> believed to be clean.
--
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://listserv.lsstcorp.org/pipermail/dm-users/attachments/20140805/634196f2/attachment.html>
More information about the dm-users
mailing list