[MAINT] remove obsolete version comparison

nilearn/_utils/testing.py

@@ -10,9 +10,6 @@
 from pathlib import Path
 
 import pytest
-import sklearn
-
-from nilearn._utils import _compare_version
 
 # we use memory_profiler library for memory consumption checks
 try:
@@ -56,15 +53,8 @@
 
     @functools.wraps(func)
     def wrapped(*args, **kwargs):
-        # --- CAN BE REMOVED --- #
-        if _compare_version(sklearn.__version__, "<", "0.22"):
-            with pytest.deprecated_call():
-                result = func(*args, **kwargs)
-        # --- CAN BE REMOVED --- #
-
-        else:
-            with pytest.warns(FutureWarning, match=match):
-                result = func(*args, **kwargs)
+        with pytest.warns(FutureWarning, match=match):
+            result = func(*args, **kwargs)
         return result
 
     return wrapped

nilearn/image/resampling.py

@@ -8,12 +8,9 @@
 import warnings
 
 import numpy as np
-import scipy
 from scipy import linalg
 from scipy.ndimage import affine_transform, find_objects
 
-from nilearn._utils import _compare_version
-
 from .. import _utils
 from .._utils import stringify_path
 from .._utils.niimg import _get_data
@@ -300,8 +297,9 @@ def _resample_one_img(
 
     # Suppresses warnings in https://github.com/nilearn/nilearn/issues/1363
     with warnings.catch_warnings():
-        if _compare_version(scipy.__version__, ">=", "0.18"):
-            warnings.simplefilter("ignore", UserWarning)
+        warnings.filterwarnings(
+            "ignore", message=".*has changed in SciPy 0.18.*"
+        )
         # The resampling itself
         affine_transform(
             data,
@@ -316,8 +314,9 @@ def _resample_one_img(
     if has_not_finite:
         # Suppresses warnings in https://github.com/nilearn/nilearn/issues/1363
         with warnings.catch_warnings():
-            if _compare_version(scipy.__version__, ">=", "0.18"):
-                warnings.simplefilter("ignore", UserWarning)
+            warnings.filterwarnings(
+                "ignore", message=".*has changed in SciPy 0.18.*"
+            )
             # We need to resample the mask of not_finite values
             not_finite = affine_transform(
                 not_finite,

nilearn/plotting/displays/_slicers.py

@@ -1,14 +1,13 @@
 import collections
 import numbers
 
-import matplotlib
 import matplotlib.pyplot as plt
 import numpy as np
 from matplotlib.colorbar import ColorbarBase
 from matplotlib.colors import LinearSegmentedColormap, ListedColormap
 from matplotlib.transforms import Bbox
 
-from nilearn._utils import _compare_version, check_niimg_3d
+from nilearn._utils import check_niimg_3d
 from nilearn._utils.docs import fill_doc
 from nilearn._utils.niimg import _is_binary_niimg, _safe_get_data
 from nilearn.image import get_data, new_img_like, reorder_img
@@ -462,13 +461,9 @@ def _map_show(
         ims = []
         to_iterate_over = zip(self.axes.values(), data_2d_list)
         for display_ax, data_2d in to_iterate_over:
-            if data_2d is not None and data_2d.min() is not np.ma.masked:
+            if data_2d is not None:
                 # If data_2d is completely masked, then there is nothing to
-                # plot. Hence, no point to do imshow(). Moreover, we see
-                # problem came up with matplotlib 2.1.0 (issue #9280) when
-                # data is completely masked or with numpy < 1.14
-                # (issue #4595). This work around can be removed when bumping
-                # matplotlib version above 2.1.0
+                # plot. Hence, no point to do imshow().
                 im = display_ax.draw_2d(
                     data_2d, data_bounds, bounding_box, type=type, **kwargs
                 )
@@ -523,10 +518,7 @@ def _show_colorbar(
             - (self._colorbar_margin["top"] + self._colorbar_margin["bottom"]),
         ]
         self._colorbar_ax = figure.add_axes(lt_wid_top_ht)
-        if _compare_version(matplotlib.__version__, ">=", "1.6"):
-            self._colorbar_ax.set_facecolor("w")
-        else:
-            self._colorbar_ax.set_axis_bgcolor("w")
+        self._colorbar_ax.set_facecolor("w")
 
         our_cmap = plt.get_cmap(cmap)
         # edge case where the data has a single value
@@ -902,10 +894,7 @@ def _init_axes(self, **kwargs):
                 [0.3 * index * (x1 - x0) + x0, y0, 0.3 * (x1 - x0), y1 - y0],
                 aspect="equal",
             )
-            if _compare_version(matplotlib.__version__, ">=", "1.6"):
-                ax.set_facecolor(facecolor)
-            else:
-                ax.set_axis_bgcolor(facecolor)
+            ax.set_facecolor(facecolor)
 
             ax.axis("off")
             coord = self.cut_coords[
@@ -1173,10 +1162,7 @@ def _init_axes(self, **kwargs):
             axes_coords = self._find_initial_axes_coord(index)
             ax = fh.add_axes(axes_coords, aspect="equal")
 
-            if _compare_version(matplotlib.__version__, ">=", "1.6"):
-                ax.set_facecolor(facecolor)
-            else:
-                ax.set_axis_bgcolor(facecolor)
+            ax.set_facecolor(facecolor)
 
             ax.axis("off")
             coord = self.cut_coords[

nilearn/plotting/glass_brain.py

@@ -3,12 +3,9 @@
 import json
 import os
 
-import matplotlib
 from matplotlib import colors, patches, transforms
 from matplotlib.path import Path
 
-from nilearn._utils import _compare_version
-
 
 def _codes_bezier(pts):
     bezier_num = len(pts)
@@ -160,10 +157,7 @@ def plot_brain_schematics(ax, direction, **kwargs):
         Useful for the caller to be able to set axes limits.
 
     """
-    if _compare_version(matplotlib.__version__, '>=', "2.0"):
-        get_axis_bg_color = ax.get_facecolor()
-    else:
-        get_axis_bg_color = ax.get_axis_bgcolor()
+    get_axis_bg_color = ax.get_facecolor()
 
     black_bg = colors.colorConverter.to_rgba(get_axis_bg_color) \
         == colors.colorConverter.to_rgba('k')

nilearn/plotting/tests/test_displays.py

@@ -317,7 +317,6 @@ def test_user_given_cmap_with_colorbar(img):
 @pytest.mark.parametrize("display", [OrthoSlicer, LYRZProjector])
 def test_data_complete_mask(display):
     """Test for a special case due to matplotlib 2.1.0.
-
     When the data is completely masked, then we have plotting issues
     See similar issue #9280 reported in matplotlib. This function
     tests the patch added for this particular issue.

nilearn/surface/surface.py

@@ -104,15 +104,16 @@ def _sample_locations_between_surfaces(
         mesh, inner_mesh, affine, n_points=10, depth=None):
     outer_vertices, _ = mesh
     inner_vertices, _ = inner_mesh
-    # when we drop support for np 1.5 replace the next 2 lines with
-    # sample_locations = np.linspace(inner_vertices, outer_vertices, n_points)
+
     if depth is None:
         steps = np.linspace(0, 1, n_points)[:, None, None]
     else:
         steps = np.asarray(depth)[:, None, None]
+
     sample_locations = outer_vertices + steps * (
         inner_vertices - outer_vertices)
     sample_locations = np.rollaxis(sample_locations, 1)
+
     sample_locations_voxel_space = np.asarray(
         resampling.coord_transform(
             *np.vstack(sample_locations).T,

nilearn/surface/tests/test_surface.py

@@ -562,24 +562,30 @@ def test_sample_locations_depth(depth, n_points):
 
 @pytest.mark.parametrize(
     "depth,n_points",
-    [(None, 1), (None, 7), ([0.], 8), ([-1.], 8),
-     ([1.], 8), ([-1., 0., .5], 8)])
+    [(None, 1),
+     (None, 7),
+     ([0.], 8),
+     ([-1.], 8),
+     ([1.], 8),
+     ([-1., 0., .5], 8)])
 def test_sample_locations_between_surfaces(depth, n_points):
     inner = flat_mesh(5, 7)
     outer = inner[0] + [0., 0., 1.], inner[1]
+
     locations = surface._sample_locations_between_surfaces(
         outer, inner, np.eye(4), n_points=n_points, depth=depth)
+
     if depth is None:
-        # can be simplified when we drop support for np 1.15
-        # (broadcasting linspace)
         expected = np.asarray(
             [np.linspace(b, a, n_points)
              for (a, b) in zip(inner[0].ravel(), outer[0].ravel())])
         expected = np.rollaxis(
             expected.reshape((*outer[0].shape, n_points)), 2, 1)
+
     else:
         offsets = ([[0., 0., - z] for z in depth])
         expected = np.asarray([vertex + offsets for vertex in outer[0]])
+
     assert np.allclose(locations, expected)