Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

MulticoreTSNE on python 3.7.3 using conda #67

Open
neurOnur opened this issue Jun 24, 2019 · 3 comments
Open

MulticoreTSNE on python 3.7.3 using conda #67

neurOnur opened this issue Jun 24, 2019 · 3 comments

Comments

@neurOnur
Copy link

I am using a python 3.7 version and could not install MulticoreTSNE using conda, or using pip (pip install MulticoreTSNE) since it tries to downgrade a few installed packages, including python itself (to 3.6.8). Below is the error message.

Should we be expecting MulticoreTSNE to be compatable with python 3.7, or would you recommend installing 3.6.8? I would avoid the latter, since it means quite a bit of reloading.

Cheers

Upon trying to install:

The following packages will be DOWNGRADED:

_ipyw_jlab_nb_ext~ 0.1.0-py37_0 --> 0.1.0-py36_0
louvain 0.6.1-py37h0a44026_2 --> 0.6.1-py36h0a44026_2
mkl-service 1.1.2-py37hfbe908c_5 --> 1.1.2-py36hfbe908c_5
navigator-updater 0.2.1-py37_0 --> 0.2.1-py36_0
pot 0.5.1-py37h1702cab_1000 --> 0.5.1-py36h1702cab_1000
pycairo 1.18.0-py37ha54c0a8_1000 --> 1.18.0-py36ha54c0a8_1000
pycurl 7.43.0.2-py37ha12b0ac_0 --> 7.43.0.2-py36ha12b0ac_0
pyqt 5.9.2-py37h655552a_2 --> 5.9.2-py36h655552a_2
pyreadr 0.1.9-py37h2573ce8_0 --> 0.1.9-py36h2573ce8_0
python 3.7.3-h359304d_0 --> 3.6.8-haf84260_0
python-igraph 0.7.1.post7-py37h01d97ff_0 --> 0.7.1.post7-py36h01d97ff_0
sphinxcontrib 1.0-py37_1 --> 1.0-py36_1

@zdwhite
Copy link

zdwhite commented Jun 24, 2019

I was able to get multicore tsne on python 3.7 if you're running it on windows, you'll need to follow the instructions here in order to get it working.

@neurOnur
Copy link
Author

Sorry for the radio silence; I have been trying to find out what is happening. In the end I feel this is not a problem with multicoreTSNE itself but other software I need for its installation.

I am a Mac user, and also tried to install MulticoreTSNE on linux based hpc. In both, I get an error message (long) which makes me think that something is wrong with my cmake or similar. I can use it to install other packages that require cmake, and the version seems to be alright

ending with either of the following:

$ pip install multicoreTSNE
or
$ git clone https://github.com/DmitryUlyanov/Multicore-TSNE.git
$ cd Multicore-TSNE/
$ pip install .

ERROR: Command "/anaconda3/bin/python -u -c 'import setuptools, tokenize;file='"'"'/private/var/folders/qq/sgxv2d656r3c49k_bp0l696w0000gn/T/pip-install-tobo8p_u/multicoreTSNE/setup.py'"'"';f=getattr(tokenize, '"'"'open'"'"', open)(file);code=f.read().replace('"'"'\r\n'"'"', '"'"'\n'"'"');f.close();exec(compile(code, file, '"'"'exec'"'"'))' install --record /private/var/folders/qq/sgxv2d656r3c49k_bp0l696w0000gn/T/pip-record-uux_cqpb/install-record.txt --single-version-externally-managed --compile" failed with error code 1 in /private/var/folders/qq/sgxv2d656r3c49k_bp0l696w0000gn/T/pip-install-tobo8p_u/multicoreTSNE/

and includes

_cmake version 3.13.3

CMake suite maintained and supported by Kitware (kitware.com/cmake).
-- The CXX compiler identification is Clang 4.0.1
-- Check for working CXX compiler: /anaconda3/bin/x86_64-apple-darwin13.4.0-clang++
-- Check for working CXX compiler: /anaconda3/bin/x86_64-apple-darwin13.4.0-clang++ -- works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Could NOT find OpenMP_CXX (missing: OpenMP_CXX_FLAGS OpenMP_CXX_LIB_NAMES) (Required is at least version "2")
-- Could NOT find OpenMP (missing: OpenMP_CXX_FOUND) (Required is at least version "2")
CMake Warning at CMakeLists.txt:22 (MESSAGE):
  Not using OpenMP.  Performance will suffer._

When I use
$ python setup.py install
The result is a bit different, ending with

_clang-4.0: error: linker command failed with exit code 1 (use -v to see invocation)
make[2]: *** [CMakeFiles/tsne_multicore.dir/build.make:102: /Users/onurbasak/Documents/2_Hpc_sequencing/software/Multicore-TSNE/build/lib.macosx-10.9-x86_64-3.7/MulticoreTSNE/libtsne_multicore.so] Error 1
make[2]: Leaving directory '/Users/onurbasak/Documents/2_Hpc_sequencing/software/Multicore-TSNE/build/temp.macosx-10.9-x86_64-3.7'
make[1]: *** [CMakeFiles/Makefile2:76: CMakeFiles/tsne_multicore.dir/all] Error 2
make[1]: Leaving directory '/Users/onurbasak/Documents/2_Hpc_sequencing/software/Multicore-TSNE/build/temp.macosx-10.9-x86_64-3.7'
make: *** [Makefile:87: all] Error 2

ERROR: Cannot find make? See above errors._

I am sure I have cmake, clang etc. but there might be a path problem... I will go through this with a specialist

I also tried the conda version, but couldnt make it work. The version of python for this seems to be <3.7.0:

$ conda search multicore-tsne --info

...
_multicore-tsne 0.1_d4ff4aab py36h3e44d54_0

file name : multicore-tsne-0.1_d4ff4aab-py36h3e44d54_0.tar.bz2
name : multicore-tsne
version : 0.1_d4ff4aab
build : py36h3e44d54_0
build number: 0
size : 70 KB
license : BSD-3-Clause
subdir : linux-64
url : https://conda.anaconda.org/conda-forge/linux-64/multicore-tsne-0.1_d4ff4aab-py36h3e44d54_0.tar.bz2
md5 : f14137001befebcea0988f12633f1dda
timestamp : 2018-11-12 17:04:10 UTC
dependencies:

  • cffi >=1.11.5
  • libstdcxx-ng >=4.9
  • numpy >=1.9.3,<2.0a0
  • python >=3.6,<3.7.0a0_

$ python --version
Python 3.7.3

@neurOnur
Copy link
Author

neurOnur commented Aug 6, 2019

Ok, the issue is solved.

My conda crushed during an update, which lead me to reinstall miniconda and mange my environments a bit better. I could install MulticoreTSNE for python 3.7.3 following 'git clone...'. The conda version (conda install -c conda-forge multicore-tsne) installed in 3.6.8 my failed in 3.7.3 in my hands.

Overall this was either due to the presence of too many packages in the env, or something that was corrupted (or paths lost) on the way.

Sorry to bother with a none multicoreTSNE related issue

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants