format with black

Author: pratikrathore8

deepxde/model.py

@@ -658,8 +658,10 @@ def train(
                 elif self.opt_name == "NNCG":
                     self._train_pytorch_nncg(iterations, display_every)
                 else:
-                    raise ValueError("Only 'L-BFGS' and 'NNCG' are supported as \
-                                      external optimizers for PyTorch.")
+                    raise ValueError(
+                        "Only 'L-BFGS' and 'NNCG' are supported as \
+                                      external optimizers for PyTorch."
+                    )
             elif backend_name == "paddle":
                 self._train_paddle_lbfgs()
         else:
@@ -827,7 +829,6 @@ def _train_pytorch_nncg(self, iterations, display_every):
             if self.stop_training:
                 break
 
-
     def _train_paddle_lbfgs(self):
         prev_n_iter = 0
 

deepxde/optimizers/config.py

@@ -60,6 +60,7 @@ def set_LBFGS_options(
     LBFGS_options["maxfun"] = maxfun if maxfun is not None else int(maxiter * 1.25)
     LBFGS_options["maxls"] = maxls
 
+
 def set_NNCG_options(
     lr=1,
     rank=10,
@@ -69,7 +70,7 @@ def set_NNCG_options(
     cgtol=1e-16,
     cgmaxiter=1000,
     lsfun="armijo",
-    verbose=False
+    verbose=False,
 ):
     """Sets the hyperparameters of NysNewtonCG (NNCG).
 
@@ -80,11 +81,11 @@ def set_NNCG_options(
             Rank of preconditioner matrix used in preconditioned conjugate gradient.
         mu (float): `mu` (torch).
             Hessian damping parameter.
-        updatefreq (int): How often the preconditioner matrix in preconditioned 
+        updatefreq (int): How often the preconditioner matrix in preconditioned
             conjugate gradient is updated. This parameter is not directly used in NNCG,
             instead it is used in _train_pytorch_nncg in deepxde/model.py.
         chunksz (int): `chunk_size` (torch).
-            Number of Hessian-vector products to compute in parallel when constructing 
+            Number of Hessian-vector products to compute in parallel when constructing
             preconditioner. If `chunk_size` is 1, the Hessian-vector products are
             computed serially.
         cgtol (float): `cg_tol` (torch).
@@ -110,6 +111,7 @@ def set_NNCG_options(
     NNCG_options["lsfun"] = lsfun
     NNCG_options["verbose"] = verbose
 
+
 def set_hvd_opt_options(
     compression=None,
     op=None,

deepxde/optimizers/pytorch/nncg.py

@@ -81,7 +81,7 @@ class NNCG(Optimizer):
 
     The parameters rank and mu will probably need to be tuned for your specific problem.
     If the optimizer is running very slowly, you can try one of the following:
-    - Increase the rank (this should increase the 
+    - Increase the rank (this should increase the
     accuracy of the Nyström approximation in PCG)
     - Reduce cg_tol (this will allow PCG to terminate with a less accurate solution)
     - Reduce cg_max_iters (this will allow PCG to terminate after fewer iterations)
@@ -155,7 +155,7 @@ def step(self, closure):
 
         Args:
             closure (callable): A closure that reevaluates the model
-              and returns the loss w.r.t. the parameters. 
+              and returns the loss w.r.t. the parameters.
         """
         if self.n_iters == 0:
             # Store the previous direction for warm starting PCG

examples/pinn_forward/Burgers_NNCG.py

@@ -1,4 +1,5 @@
 """Backend supported: tensorflow.compat.v1, tensorflow, pytorch, paddle"""
+
 import deepxde as dde
 import numpy as np
 
@@ -49,7 +50,10 @@ def pde(x, y):
 y_pred = model.predict(X)
 f = model.predict(X, operator=pde)
 print("Mean residual after Adam+L-BFGS:", np.mean(np.absolute(f)))
-print("L2 relative error after Adam+L-BFGS:", dde.metrics.l2_relative_error(y_true, y_pred))
+print(
+    "L2 relative error after Adam+L-BFGS:",
+    dde.metrics.l2_relative_error(y_true, y_pred),
+)
 np.savetxt("test_adam_lbfgs.dat", np.hstack((X, y_true, y_pred)))
 
 # Run NNCG after Adam+L-BFGS
@@ -62,6 +66,8 @@ def pde(x, y):
 y_pred = model.predict(X)
 f = model.predict(X, operator=pde)
 print("Mean residual after Adam+L-BFGS+NNCG:", np.mean(np.absolute(f)))
-print("L2 relative error after Adam+L-BFGS+NNCG:",
-      dde.metrics.l2_relative_error(y_true, y_pred))
+print(
+    "L2 relative error after Adam+L-BFGS+NNCG:",
+    dde.metrics.l2_relative_error(y_true, y_pred),
+)
 np.savetxt("test_adam_lbfgs_nncg.dat", np.hstack((X, y_true, y_pred)))