update docs

flexsolve/__init__.py

@@ -34,4 +34,4 @@
 from .problem import *
 from .profiler import *
 
-__version__ = '0.4.2'
+__version__ = '0.4.3'

flexsolve/utils.py

@@ -53,14 +53,14 @@ def aitken_iter(x, gg, dxg, dgg_g):
         return scalar_aitken_iter(x, gg, dxg, dgg_g)
 
 @overload(wegstein_iter)
-def jit_wegstein_iter(x, dx, g1, g0):
+def jit_wegstein_iter(x, dx, g1, g0):  # pragma: no cover
     if isinstance(x, types.Array) and x.ndim:
         return array_wegstein_iter  
     else:
         return scalar_wegstein_iter
 
 @overload(aitken_iter)
-def jit_aitken_iter(x, gg, dxg, dgg_g):
+def jit_aitken_iter(x, gg, dxg, dgg_g):  # pragma: no cover
     if isinstance(x, types.Array) and x.ndim:
         return array_aitken_iter
     else:
@@ -147,15 +147,15 @@ def IQ_iter(y0, y1, y2, x0, x1, x2, dx, df0, xlast):
     return x
 
 @njitable(cache=True)
-def raise_iter_error(): # pragma: no cover
+def raise_iter_error():
     raise RuntimeError('maximum number of iterations exceeded; root could not be solved')
 
 @njitable(cache=True)
 def raise_tol_error(): # pragma: no cover
     raise RuntimeError('minimum tolerance reached; root could not be solved')
 
 @njitable(cache=True)
-def raise_convergence_error(): # pragma: no cover
+def raise_convergence_error():
     raise RuntimeError('objective function either oscillates or diverges from solution; root could not be solved')
 
 @njitable(cache=True)
@@ -183,7 +183,7 @@ def fixedpoint_converged(dx, xtol):
         return scalar_fixedpoint_converged(dx, xtol)
 
 @overload(fixedpoint_converged)
-def jit_fixedpoint_converged(dx, xtol):
+def jit_fixedpoint_converged(dx, xtol): # pragma: no cover
     if isinstance(dx, types.Array) and dx.ndim:
         return array_fixedpoint_converged
     else:
@@ -196,5 +196,5 @@ def mean(x):
 def scalar_mean(x): return x
 
 @overload(mean)
-def jit_mean(x):
+def jit_mean(x): # pragma: no cover
     return np.mean if isinstance(x, types.Array) and x.ndim else scalar_mean

tests/test_fixedpoint_array_solvers.py

@@ -78,11 +78,6 @@ def test_fixedpoint_array_solvers():
 
     assert p.sizes() == {'Wegstein': 5, 'Aitken': 5, 'Lstsq': 22, 'Fixed point': 194}
 
-    flx.speed_up()
-    solution = flx.wegstein(f, feed, convergenceiter=4, xtol=5e-8)
-    assert_allclose(solution, real_solution)
-    solution = flx.aitken(f, feed, convergenceiter=4, xtol=5e-8)
-    assert_allclose(solution, real_solution)
 
 def test_fixedpoint_array_solvers2():
     original_feed = feed.copy()
@@ -95,6 +90,8 @@ def test_fixedpoint_array_solvers2():
         solution = flx.wegstein(f2, feed, convergenceiter=4, xtol=1e-8, maxiter=200)
     with pytest.raises(RuntimeError):
         solution = flx.wegstein(f2, feed, xtol=1e-8, maxiter=20)
+    with pytest.raises(RuntimeError):
+        solution = flx.aitken(f2, feed, convergenceiter=4, xtol=1e-8, maxiter=200)
 
     solution = flx.wegstein(p, feed, checkconvergence=False, convergenceiter=4, xtol=1e-8)
     p.archive('Wegstein early termination')
@@ -135,6 +132,13 @@ def test_fixedpoint_array_solvers2():
                          'Lstsq': 27, 'Lstsq early termination': 29, 
                          'Fixed point': 191, 'Fixed point early termination': 191}
 
+def test_fixedpoint_array_with_speed_up():
+    flx.speed_up()
+    solution = flx.wegstein(f, feed, convergenceiter=4, xtol=5e-8)
+    assert_allclose(solution, real_solution)
+    solution = flx.aitken(f, feed, convergenceiter=4, xtol=5e-8)
+    assert_allclose(solution, real_solution)
+
 def test_conditional_fixedpoint_array_solvers():
     original_feed = feed.copy()