Package Details: mlpack 3.1.1-2

Git Clone URL: https://aur.archlinux.org/mlpack.git (read-only)
Package Base: mlpack
Description: a scalable c++ machine learning library
Upstream URL: http://www.mlpack.org
Licenses: BSD
Submitter: jerry73204
Maintainer: jerry73204
Last Packager: jerry73204
Votes: 7
Popularity: 0.189104
First Submitted: 2015-10-29 03:40
Last Updated: 2019-06-05 15:13

Latest Comments

1 2 Next › Last »

nicolino commented on 2019-07-16 17:47

I have had the same problems as Relyak. I have modified the PKGBUILD as indicated. The lib64 conflict was resolved. The others did not.

Reylak commented on 2019-06-14 10:46

I second masca's comment and its conflict errors. Mine are:

mlpack: /usr/bin/cygdb exists in filesystem (owned by cython)
mlpack: /usr/bin/cython exists in filesystem (owned by cython)
mlpack: /usr/bin/cythonize exists in filesystem (owned by cython)
mlpack: /usr/bin/f2py exists in filesystem (owned by python-numpy)
mlpack: /usr/bin/f2py3 exists in filesystem (owned by python-numpy)
mlpack: /usr/bin/f2py3.7 exists in filesystem (owned by python-numpy)
mlpack: /usr/lib64 exists in filesystem (owned by filesystem)

(same as masca's but without the "easy-install.pth" one; I see in the PKGBUILD that "easy-install.pth" is deleted in the package() function)

For the lib64 conflict, it seems that CMake does not properly handle the UsrMerge of ArchLinux with your PKGBUILD. Passing -D CMAKE_INSTALL_LIBDIR="lib" to cmake fixes the issue as suggested here. This is what is used by this package for example. That does not solve the other conflicts though.

EDIT: the other conflicts actually ended up disappearing but I'm not sure why, sorry :(

masca commented on 2019-06-05 15:26

I am having troubles installing. I tried on two different machines and I get errors (albeit different ones). In the first case, I get the following:

build/temp.linux-x86_64-3.6/home/andrea/Documenti/mlpack/src/mlpack-3.1.1/src/mlpack/bindings/python/mlpack/sparse_coding.o: file not recognized: file format not recognized collect2: error: ld returned 1 exit status error: command 'g++' failed with exit status 1 make[2]: [src/mlpack/methods/sparse_coding/CMakeFiles/build_pyx_sparse_coding.dir/build.make:58: src/mlpack/methods/sparse_coding/CMakeFiles/build_pyx_sparse_coding] Error 1 make[1]: [CMakeFiles/Makefile2:7344: src/mlpack/methods/sparse_coding/CMakeFiles/build_pyx_sparse_coding.dir/all] Error 2 make: *** [Makefile:141: all] Error 2

whereas in the second case I get: mlpack: /usr/bin/cygdb esiste già nel filesystem mlpack: /usr/bin/cython esiste già nel filesystem mlpack: /usr/bin/cythonize esiste già nel filesystem mlpack: /usr/bin/f2py esiste già nel filesystem mlpack: /usr/bin/f2py3 esiste già nel filesystem mlpack: /usr/bin/f2py3.7 esiste già nel filesystem mlpack: /usr/lib/python3.7/site-packages/easy-install.pth esiste già nel filesystem mlpack: /usr/lib64 esiste già nel filesystem

jerry73204 commented on 2019-06-05 15:14

Hi @mark, I drop the version constraint for armadillo. Hope it works for you.

mark commented on 2019-01-21 10:01

@jerry73204 - it compiled successfully, thanks!

Also, it still complains about armadillo>=6.500.0, that it cannot be found so I have to install it explicitly. Other than that it's great!

jerry73204 commented on 2019-01-16 07:03

I update the package to 3.0.4 and disable check by default.

Mark please check if this update fixes your problem.

mark commented on 2019-01-15 08:43

gcc: error: unrecognized command line option ‘-std=c++11-march=x86-64’ error: command 'gcc' failed with exit status 1 make[2]: *** [src/mlpack/methods/sparse_coding/CMakeFiles/build_pyx_sparse_coding.dir/build.make:58: src/mlpack/methods/sparse_coding/CMakeFiles/build_pyx_sparse_coding] Error 1

This is when installing it with trizen, first it complained armadillo>=6.500.0 cannot be found, but it went away after installing it manually.

Is this a typo in the makefile or? Should there be a whitespace between 11 and the dash?

alcros33 commented on 2018-05-14 16:31

to jerry73204,

Thank you, I'll just skip the check() step, but just saying that I have reproduced the issue around 10 times...

jerry73204 commented on 2018-05-12 18:16

To alcron33, My build runs successfully and I cannot reproduce this error. The message indicates that it was caused by a failing test. It's better to report it to upstream. If you need the build to be done, as a remedy, you can comment out check() function in PKGBUILD.

alcros33 commented on 2018-05-11 18:29

I have tried installing a couple of times... it always displays this error when doing the check() after build:

Running 1262 test cases... /tmp/yaourt-tmp-alcros33/aur-mlpack/src/mlpack-3.0.0/src/mlpack/tests/distribution_test.cpp(888): fatal error: in "DistributionTest/GammaDistributionRandomTest": difference{0.100615} between d2.Beta(i){1.1811578593978589} and b(i){1.3} exceeds 10%

unknown location(0): fatal error: in "DecisionTreeMainTest/DecisionTreeCategoricalOutputDimensionTest": memory access violation at address: 0x00000000: no mapping at fault address

/tmp/yaourt-tmp-alcros33/aur-mlpack/src/mlpack-3.0.0/src/mlpack/tests/main_tests/decision_tree_test.cpp(98): last checkpoint: "DecisionTreeCategoricalOutputDimensionTest" test entry

*** 2 failures are detected in the test module "mlpackTest"