[MOD] - modified Lorenz function to stabilize gpc test

pygpc/testfunctions/testfunctions.py

@@ -4259,7 +4259,7 @@ def lorenz(t, state, sigma, beta, rho):
             y0 = [self.p["x_0"][i], self.p["y_0"][i], self.p["z_0"][i]]
             # only save x-coordinate (index 0)
             sols[i, :] = odeint(lorenz, y0, t, p, tfirst=True)[:, 0]
-        x_out = sols
+        x_out = sols[:, 1:]   # skip first timestep
 
         return x_out
 
@@ -4330,6 +4330,6 @@ def simulate(self, process_id=None, matlab_engine=None):
             y0 = [self.p["x_0"][i], self.p["y_0"][i], self.p["z_0"][i]]
             # only save x-coordinate (index 0)
             sols[i, :] = Main.Julia_Lorenz(p, y0, t)[0]
-        x_out = sols
+        x_out = sols[:, 1:]     # skip first timestep
 
         return x_out

tests/test_postprocessing.py

@@ -227,7 +227,7 @@ def test_postprocessing_002_plot_functions(self):
         options["n_cpu"] = 0
         options["error_type"] = 'nrmsd'
         options["error_norm"] = 'absolute'
-        options["n_samples_validation"] = 1000
+        options["n_samples_validation"] = 10
         options["matrix_ratio"] = 5
         options["fn_results"] = fn_results
         options["eps"] = 0.01
@@ -259,7 +259,7 @@ def test_postprocessing_002_plot_functions(self):
                                               fn_out=fn_results + "_sens_summary.txt")
 
         # plot time course of mean together with probability density, sobol sensitivity coefficients and global derivatives
-        t = np.arange(0.0, parameters["t_end"], parameters["step_size"])
+        t = np.arange(parameters["step_size"], parameters["t_end"], parameters["step_size"])
         pygpc.plot_sens_summary(session=session,
                                 coeffs=coeffs,
                                 sobol=sobol,