starting test cleanup

Author: pancetta

pySDC/playgrounds/Gander/matrix_based.py

@@ -118,5 +118,5 @@ def estimates_over_lambda():
     plt.show()
 
 if __name__ == '__main__':
-    iteration_vs_estimate()
-    # estimates_over_lambda()
+    # iteration_vs_estimate()
+    estimates_over_lambda()

pySDC/playgrounds/Gander/two_grid.py

@@ -0,0 +1,100 @@
+import numpy as np
+import scipy.sparse as sp
+import scipy.sparse.linalg as spl
+import matplotlib.pyplot as plt
+
+import pySDC.helpers.transfer_helper as th
+
+np.random.seed(32)
+
+nX = 2 ** 3 - 1
+nXc = 2 ** 2 - 1
+dx = 1/(nX + 1)
+dxc = 1/(nXc + 1)
+
+stencil = [1, -2, 1]
+A = sp.diags(stencil, [-1, 0, 1], shape=(nX, nX), format='csc')
+A *= - 1.0 / (dx ** 2)
+b = np.zeros(nX)
+# %%
+PGS = spl.inv(sp.csc_matrix(np.tril(A.todense())))
+PJ = sp.dia_matrix(np.diag(1/np.diag(A.todense())))
+
+RGS = sp.eye(nX) - PGS @ A
+RJ = sp.eye(nX) - PJ @ A
+sRJ = sp.eye(nX) - 0.5 * PJ @ A
+
+Ac = sp.diags(stencil, [-1, 0, 1], shape=(nX//2, nX//2), format='csc')
+Ac *= - 1.0 / (dxc ** 2)
+Acinv = spl.inv(Ac)
+PJc = sp.dia_matrix(np.diag(1/np.diag(Ac.todense())))
+
+fine_grid = np.linspace(dx, 1, nX, endpoint=False)
+coarse_grid = np.linspace(dxc, 1, nXc, endpoint=False)
+Pr = th.interpolation_matrix_1d(fine_grid, coarse_grid, k=2, periodic=False, equidist_nested=True)
+Prinj = th.interpolation_matrix_1d(fine_grid, coarse_grid, k=0, periodic=False, equidist_nested=True)
+Re = 0.5 * Pr.T
+# Re = Prinj.T
+
+TG = (sp.eye(nX) - Pr @ Acinv @ Re @ A) @ sRJ.power(3)
+TJ = (sp.eye(nX) - Pr @ PJc @ Re @ A) @ sRJ.power(3)
+
+# %%
+
+nIter = 100
+dK = 10
+
+def coarse(x, dK=dK):
+    xK = x
+    for k in range(dK):
+        xK = RJ @ xK #+ RJ @ b
+    return xK
+
+
+def fine(x, dK=dK):
+    xK = x
+    for k in range(dK):
+        xK = TG @ xK #+ PGS @ b
+    return xK
+
+def initBlocks(x):
+    x0 = np.sin(np.linspace(dx, 1*2*np.pi, nX))
+    # x0 = np.random.rand(nX)
+    for l in range(nB+1):
+        x[l, :] = x0
+
+nB = nIter//dK
+nK = nB+1
+
+
+uSeq = np.zeros((nB+1, nX))
+uNum = np.zeros((nK+1, nB+1, nX))
+
+initBlocks(uNum[0])
+
+uSeq[0] = uNum[0, 0]
+uNum[:, 0, :] = uNum[0, 0, :]
+
+for k in range(nK):
+    for l in range(nB):
+
+        uF = fine(uNum[k, l])
+        uGk = coarse(uNum[k, l], dK=1)
+        uGkp1 = coarse(uNum[k+1, l], dK=1)
+
+        uNum[k+1, l+1] = uF + 1*(uGkp1 - uGk)
+
+
+for l in range(nB):
+    uSeq[l+1] = fine(uSeq[l])
+
+# %%
+
+err = np.linalg.norm(uNum, axis=-1)[:, -1]
+errSeq = np.linalg.norm(uSeq, axis=-1)
+
+plt.figure()
+plt.semilogy(err, label='Parareal')
+plt.semilogy(errSeq, label='Sequential')
+plt.legend()
+plt.show()

pySDC/projects/soft_failure/visualization_helper.py

@@ -31,7 +31,7 @@ def show_residual_across_simulation(stats, fname):
     residual = np.zeros(maxiter)
     residual[:] = -99
     for k, v in extract_stats.items():
-        if k.iter is not -1:
+        if k.iter != -1:
             residual[k.iter - 1] = np.log10(v)
 
     # Set up latex stuff and fonts

pySDC/tests/test_tutorials/test_step_7.py

@@ -1,11 +1,12 @@
 import os
 import subprocess
+import pytest
 
-from pySDC.tutorial.step_7.A_pySDC_with_FEniCS import main as main_A
 from pySDC.tutorial.step_7.B_pySDC_with_mpi4pyfft import main as main_B
 
-
+@pytest.mark.fenics
 def test_A():
+    from pySDC.tutorial.step_7.A_pySDC_with_FEniCS import main as main_A
     main_A()