forked from lisa-lab/pylearn2
-
Notifications
You must be signed in to change notification settings - Fork 0
/
.travis.yml
71 lines (63 loc) · 3.6 KB
/
.travis.yml
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
# After changing this file, check it on:
# http://lint.travis-ci.org/
## Use language: c to avoid being trapped in a virtualenv,
# so we can use "apt-get" (instead of "pip" )to install scipy,
# as pip takes much too long.
#language: c
language: python
python:
#Currently we use functools.cmp_to_key that is available only in python 2.7
- "2.6"
# command to install dependencies
before_install:
#zlib1g-dev is needed to allow PIL to uncompress the dataset.
- sudo apt-get install -qq libatlas3gf-base libatlas-dev liblapack-dev gfortran zlib1g-dev
libhdf5-serial-dev
# The "|| cat ..." part prints the error log if the pip command fails
install:
#Make sure the pip version used is linked to the used python version
- "cat `which pip`"
- "which python"
- "python --version"
# If we don't install numpy before SciPy, the SciPy installations fails.
# We can't compile scipy 0.9 sparse stuff with NumPy 1.5/1.5.1 due to a bug.
- "pip install -q numpy==1.6 --use-mirrors"
# We support scipy 0.7.2, but it is not available on pypi anymore.
# So we test with 0.9. Our internal buildbot have 0.7.2.
# We install it later only for the PART that need it.
- "pip install -q nose==1.0.0 --use-mirrors || (cat /home/travis/.pip/pip.log; false)"
- "pip install -q Pillow --use-mirrors || (cat /home/travis/.pip/pip.log; false)"
#numexpr 2.0 need NumPy 1.6 or above. As we support NumPy 1.5,
#we need an older numexpr version. numexpr is needed for pytables
- "pip install -q numexpr==1.4.2 --use-mirrors || (cat /home/travis/.pip/pip.log; false)"
#cython is needed by pylearn2/utils/setup.py
- "pip install -q cython --use-mirrors || (cat /home/travis/.pip/pip.log; false)"
- "pip install -q tables==2.4.0 --use-mirrors || (cat /home/travis/.pip/pip.log; false)"
- "pip install -q pyyaml --use-mirrors || (cat /home/travis/.pip/pip.log; false)"
- "pip install -q sphinx --use-mirrors || (cat /home/travis/.pip/pip.log; false)"
- "pip install git+git://github.com/numpy/numpydoc.git@e433d273afe278162b07535f54ab182bff4f2565 || (cat /home/travis/.pip/pip.log; false)"
- "pip install scipy==0.8 --use-mirrors || (cat /home/travis/.pip/pip.log; false)"
- "pip install argparse --use-mirrors || (cat /home/travis/.pip/pip.log; false)"
#More recent version then the one in the os
- "pip install -q pyflakes --use-mirrors || (cat /home/travis/.pip/pip.log; false)"
- "pip install -q git+git://git.assembla.com/jobman.git || (cat /home/travis/.pip/pip.log; false)"
- "pip install -q --no-deps git+git://github.com/Theano/Theano.git || (cat /home/travis/.pip/pip.log ; false)"
- cd pylearn2/utils/
# Change this once we have a unified setup file.
- python setup.py build_ext --inplace
- cd ../../
# command to run tests
env:
- FLAGS=floatX=float64 PART="-e test_autoencoder.py -e test_dbm.py -e test_dropout.py -e test_maxout.py -e test_mlp.py -e test_mnd.py -e test_rbm.py -e test_reflection_clip.py -e test_s3c_inference.py -e test_s3c_misc.py -e test_svm.py"
- FLAGS=floatX=float64 PART=pylearn2/models
#Exclude file inside pylearn2/models, sadly I didn't found a better way.
- FLAGS=floatX=float32 PART="-e test_autoencoder.py -e test_dbm.py -e test_dropout.py -e test_maxout.py -e test_mlp.py -e test_mnd.py -e test_rbm.py -e test_reflection_clip.py -e test_s3c_inference.py -e test_s3c_misc.py -e test_svm.py"
- FLAGS=floatX=float32 PART=pylearn2/models
- TEST_DOC=YES
# Start testing
script:
#We can't build the test dataset as the original is not present. We can't download the original as it
# is too big to download each time. If present run: python make_dataset.py
- sh .travis.sh
after_failure:
- cat /home/travis/.pip/pip.log