More standardization of floating point slop in mpl_toolkits.

anntzer · anntzer · commit c5e02450dc9a · 2022-01-27T11:08:01.000+01:00
A few instances I missed in fb18d95. At least grepping for delta doesn't yield anything else.
diff --git a/lib/mpl_toolkits/axisartist/floating_axes.py b/lib/mpl_toolkits/axisartist/floating_axes.py
@@ -5,8 +5,11 @@
 # TODO :
 # see if tick_iterator method can be simplified by reusing the parent method.
 
+import functools
+
 import numpy as np
 
+import matplotlib as mpl
 from matplotlib import _api, cbook
 import matplotlib.axes as maxes
 import matplotlib.patches as mpatches
@@ -121,10 +124,11 @@ def f1():
             dd[mm] = dd2[mm] + np.pi / 2
 
             tick_to_axes = self.get_tick_transform(axes) - axes.transAxes
+            in_01 = functools.partial(
+                mpl.transforms._interval_contains_close, (0, 1))
             for x, y, d, d2, lab in zip(xx1, yy1, dd, dd2, labels):
                 c2 = tick_to_axes.transform((x, y))
-                delta = 0.00001
-                if 0-delta <= c2[0] <= 1+delta and 0-delta <= c2[1] <= 1+delta:
+                if in_01(c2[0]) and in_01(c2[1]):
                     d1, d2 = np.rad2deg([d, d2])
                     yield [x, y], d1, d2, lab
 
diff --git a/lib/mpl_toolkits/axisartist/grid_helper_curvelinear.py b/lib/mpl_toolkits/axisartist/grid_helper_curvelinear.py
@@ -1,10 +1,13 @@
 """
 An experimental support for curvilinear grid.
 """
+
+import functools
 from itertools import chain
 
 import numpy as np
 
+import matplotlib as mpl
 from matplotlib import _api
 from matplotlib.path import Path
 from matplotlib.transforms import Affine2D, IdentityTransform
@@ -231,10 +234,11 @@ def f1():
             dd[mm] = dd2[mm] + np.pi / 2
 
             tick_to_axes = self.get_tick_transform(axes) - axes.transAxes
+            in_01 = functools.partial(
+                mpl.transforms._interval_contains_close, (0, 1))
             for x, y, d, d2, lab in zip(xx1, yy1, dd, dd2, labels):
                 c2 = tick_to_axes.transform((x, y))
-                delta = 0.00001
-                if 0-delta <= c2[0] <= 1+delta and 0-delta <= c2[1] <= 1+delta:
+                if in_01(c2[0]) and in_01(c2[1]):
                     d1, d2 = np.rad2deg([d, d2])
                     yield [x, y], d1, d2, lab
 
