Merge pull request #427 from matthew-brett/test-pre-wheels

MRG: test against pre-release numpy / scipy

Use pre-release numpy / scipy wheels for testing.

.travis.yml

@@ -5,11 +5,6 @@
 # - There can't be any leading "-"s - All newlines will be removed, so use
 # ";"s
 sudo: false  # To use travis container infrastructure
-addons:
-  apt:
-    packages:
-        - libblas-dev
-        - liblapack-dev
 language: python
 cache:
   directories:
@@ -18,6 +13,10 @@ env:
     global:
         - DEPENDS="numpy scipy sympy matplotlib nibabel"
         - INSTALL_TYPE="setup"
+        - EXTRA_WHEELS="https://5cf40426d9f06eb7461d-6fe47d9331aba7cd62fc36c7196769e4.ssl.cf2.rackcdn.com"
+        - PRE_WHEELS="https://7933911d6844c6c53a7d-47bd50c35cd79bd838daf386af554a83.ssl.cf2.rackcdn.com"
+        - EXTRA_PIP_FLAGS="--find-links=$EXTRA_WHEELS"
+        - PRE_PIP_FLAGS="--pre $EXTRA_PIP_FLAGS --find-links $PRE_WHEELS"
 python:
     - 3.4
     - 3.5
@@ -38,6 +37,11 @@ matrix:
     - python: 3.4
       env:
         - NIPY_EXTERNAL_LAPACK=1
+      addons:
+        apt:
+          packages:
+              - libblas-dev
+              - liblapack-dev
     - python: 2.7
       env:
         - INSTALL_TYPE=sdist
@@ -51,6 +55,10 @@ matrix:
         - INSTALL_TYPE=requirements
         - DEPENDS=
     - python: 3.6
+    # test 3.5 against pre-release builds of everything
+    - python: 3.5
+      env:
+        - EXTRA_PIP_FLAGS="$PRE_PIP_FLAGS"
 
 before_install:
     - source tools/travis_tools.sh
@@ -62,9 +70,9 @@ before_install:
     - pip install -U pip
     - pip install nose mock  # always
     - if [ -n "$PRE_DEPENDS" ]; then
-      pip install $PRE_DEPENDS;
+      pip install $EXTRA_PIP_FLAGS $PRE_DEPENDS;
       fi
-    - pip install $DEPENDS
+    - pip install $EXTRA_PIP_FLAGS $DEPENDS
     - if [ "${COVERAGE}" == "1" ]; then
       pip install coverage;
       pip install coveralls codecov;
@@ -78,12 +86,13 @@ install:
       elif [ "$INSTALL_TYPE" == "sdist" ]; then
         python setup_egg.py egg_info  # check egg_info while we're here
         python setup_egg.py sdist
-        pip install dist/*.tar.gz
+        pip install $EXTRA_PIP_FLAGS dist/*.tar.gz
       elif [ "$INSTALL_TYPE" == "wheel" ]; then
         pip install wheel
         python setup_egg.py bdist_wheel
-        pip install dist/*.whl
+        pip install $EXTRA_PIP_FLAGS dist/*.whl
       elif [ "$INSTALL_TYPE" == "requirements" ]; then
+        pip install $EXTRA_PIP_FLAGS -r requirements.txt
         pip install -r requirements.txt
         python setup.py install
       fi

nipy/algorithms/diagnostics/tsdiffplot.py

@@ -68,11 +68,9 @@ def xmax_labels(ax, val, xlabel, ylabel):
 
     # slice plots min max mean
     ax = axes[3]
-    ax.hold(True)
     ax.plot(np.mean(scaled_slice_diff, 0), 'k')
     ax.plot(np.min(scaled_slice_diff, 0), 'b')
     ax.plot(np.max(scaled_slice_diff, 0), 'r')
-    ax.hold(False)
     xmax_labels(ax, S+1,
                 'Slice number',
                 'Max/mean/min \n slice variation')

nipy/algorithms/group/parcel_analysis.py

@@ -46,7 +46,7 @@ def _gaussian_filter(x, msk, sigma):
     x[msk] = 0.
     gx = nd.gaussian_filter(x, sigma)
     norma = 1 - nd.gaussian_filter(msk.astype(float), sigma)
-    gx[True - msk] /= norma[True - msk]
+    gx[~msk] /= norma[~msk]
     gx[msk] = 0.
     return gx
 

nipy/algorithms/segmentation/tests/test_segmentation.py

@@ -44,11 +44,11 @@ def _test_brain_seg(model, niters=3, beta=0, ngb_size=6, init_params=None,
     _check_dims(S.sigma, 1, S.mixmat.shape[0])
     # Check that probabilities are zero outside the mask and sum up to
     # one inside the mask
-    assert_almost_equal(S.ppm[True - S.mask].sum(-1).max(), 0)
+    assert_almost_equal(S.ppm[~S.mask].sum(-1).max(), 0)
     assert_almost_equal(S.ppm[S.mask].sum(-1).min(), 1)
     # Check that labels are zero outside the mask and > 1 inside the
     # mask
-    assert_almost_equal(S.label[True - S.mask].max(), 0)
+    assert_almost_equal(S.label[~S.mask].max(), 0)
     assert_almost_equal(S.label[S.mask].min(), 1)
 
 

nipy/algorithms/statistics/formula/formulae.py

@@ -734,7 +734,7 @@ def _setup_design(self):
 
         # This renaming is here principally because of the intercept.
 
-        random_offset = np.random.random_integers(low=0, high=2**30)
+        random_offset = np.random.randint(low=0, high=2**30)
 
         terms = getterms(self.mean)
 

nipy/algorithms/statistics/tests/test_intrinsic_volumes.py

@@ -39,7 +39,7 @@ def randombox(shape):
     """
     Generate a random box, returning the box and the edge lengths
     """
-    edges = [np.random.random_integers(0, shape[j], size=(2,))
+    edges = [np.random.randint(0, shape[j] + 1, size=(2,))
              for j in range(len(shape))]
 
     for j in range(len(shape)):

nipy/algorithms/tests/test_kernel_smooth.py

@@ -3,7 +3,7 @@
 """ Test for smoothing with kernels """
 from __future__ import absolute_import
 import numpy as np
-from numpy.random import random_integers as randint
+from numpy.random import randint
 
 from ... import load_image
 from ..kernel_smooth import LinearFilter, sigma2fwhm, fwhm2sigma
@@ -76,20 +76,21 @@ def test_kernel():
     tol = 0.9999
     sdtol = 1.0e-8
     for x in range(6):
-        shape = randint(30,60,(3,))
+        shape = randint(30, 60 + 1, (3,))
         # pos of delta
-        ii, jj, kk = randint(11,17, (3,))
+        ii, jj, kk = randint(11, 17 + 1, (3,))
         # random affine coordmap (diagonal and translations)
-        coordmap = AffineTransform.from_start_step('ijk', 'xyz', 
-                                          randint(5,20,(3,))*0.25,
-                                          randint(5,10,(3,))*0.5)
+        coordmap = AffineTransform.from_start_step(
+            'ijk', 'xyz',
+            randint(5, 20 + 1, (3,)) * 0.25,
+            randint(5, 10 + 1, (3,)) * 0.5)
         # delta function in 3D array
         signal = np.zeros(shape)
         signal[ii,jj,kk] = 1.
         signal = Image(signal, coordmap=coordmap)
         # A filter with coordmap, shape matched to image
-        kernel = LinearFilter(coordmap, shape, 
-                              fwhm=randint(50,100)/10.)
+        kernel = LinearFilter(coordmap, shape,
+                              fwhm=randint(50, 100 + 1) / 10.)
         # smoothed normalized 3D array
         ssignal = kernel.smooth(signal).get_data()
         ssignal[:] *= kernel.norms[kernel.normalization]

nipy/core/image/image.py

@@ -402,6 +402,7 @@ def __iter__(self):
 
     def __eq__(self, other):
         return (isinstance(other, self.__class__)
+                and self.shape == other.shape
                 and np.all(self.get_data() == other.get_data())
                 and np.all(self.affine == other.affine)
                 and (self.axes.coord_names == other.axes.coord_names))

nipy/labs/datasets/volumes/tests/test_volume_grid.py

@@ -62,7 +62,7 @@ def test_trivial_grid():
     img = VolumeGrid(data=data,
                     transform=identity,
                     )
-    x, y, z = np.random.random_integers(N, size=(3, 10)) - 1
+    x, y, z = np.random.randint(1, N + 1, size=(3, 10)) - 1
     data_ = img.values_in_world(x, y, z)
     # Check that passing in arrays with different shapes raises an error
     yield np.testing.assert_raises, ValueError, \

nipy/modalities/fmri/hemodynamic_models.py

@@ -34,7 +34,7 @@ def _gamma_difference_hrf(tr, oversampling=16, time_length=32., onset=0.,
          hrf sampling on the oversampled time grid
     """
     dt = tr / oversampling
-    time_stamps = np.linspace(0, time_length, float(time_length) / dt)
+    time_stamps = np.linspace(0, time_length, int(float(time_length) / dt))
     time_stamps -= onset / dt
     hrf = gamma.pdf(time_stamps, delay / dispersion, dt / dispersion) - \
         ratio * gamma.pdf(
@@ -177,7 +177,7 @@ def _sample_condition(exp_condition, frametimes, oversampling=16,
 
     hr_frametimes = np.linspace(frametimes.min() + min_onset,
                                 frametimes.max() * (1 + 1. / (n - 1)),
-                                n_hr)
+                                int(n_hr))
 
     # Get the condition information
     onsets, durations, values = tuple(map(np.asanyarray, exp_condition))