Change dde.maps to dde.nn (#539)

Author: anranjiao

docs/demos/pinn_forward/burgers.rar.rst

@@ -66,7 +66,7 @@ Next, we choose the network. Here, we use a fully connected neural network of de
 
 .. code-block:: python
 
-    net = dde.maps.FNN([2] + [20] * 3 + [1], "tanh", "Glorot normal")
+    net = dde.nn.FNN([2] + [20] * 3 + [1], "tanh", "Glorot normal")
 
 Now, we have the PDE problem and the network. We build a ``Model`` and choose the optimizer and learning rate:
 

docs/demos/pinn_forward/burgers.rst

@@ -66,7 +66,7 @@ Next, we choose the network. Here, we use a fully connected neural network of de
 
 .. code-block:: python
 
-    net = dde.maps.FNN([2] + [20] * 3 + [1], "tanh", "Glorot normal")
+    net = dde.nn.FNN([2] + [20] * 3 + [1], "tanh", "Glorot normal")
 
 Now, we have the PDE problem and the network. We build a ``Model`` and choose the optimizer and learning rate:
 

docs/demos/pinn_forward/diffusion.1d.exactBC.rst

@@ -78,7 +78,7 @@ Next, we choose the network. Here, we use a fully connected neural network of de
     layer_size = [2] + [32] * 3 + [1]
     activation = "tanh"
     initializer = "Glorot uniform"
-    net = dde.maps.FNN(layer_size, activation, initializer)
+    net = dde.nn.FNN(layer_size, activation, initializer)
 
 Then we construct a function that spontaneously satisfies both the initial and the boundary conditions to transform the network output. In this case, :math:`t(1-x^2)y + sin(\pi x)` is used. When :math:`t` is equal to 0, the initial condition :math:`sin(\pi x)` is recovered. When :math:`x` is equal to -1 or 1, the boundary condition :math:`y(-1, t) = y(1, t) = 0` is recovered. Hence the initial and boundary conditions are both hard conditions.
 

docs/demos/pinn_forward/diffusion.1d.resample.rst

@@ -95,7 +95,7 @@ Next, we choose the network. Here, we use a fully connected neural network of de
     layer_size = [2] + [32] * 3 + [1]
     activation = "tanh"
     initializer = "Glorot uniform"
-    net = dde.maps.FNN(layer_size, activation, initializer)
+    net = dde.nn.FNN(layer_size, activation, initializer)
 
 The following code is to apply mini-batch gradient descent sampling method. The period is the period of resamping. Here, the training points in the domain will be resampled every 100 iterations.
 

docs/demos/pinn_forward/diffusion.1d.rst

@@ -94,7 +94,7 @@ Next, we choose the network. Here, we use a fully connected neural network of de
     layer_size = [2] + [32] * 3 + [1]
     activation = "tanh"
     initializer = "Glorot uniform"
-    net = dde.maps.FNN(layer_size, activation, initializer)
+    net = dde.nn.FNN(layer_size, activation, initializer)
 
 Now, we have the PDE problem and the network. We build a ``Model`` and choose the optimizer and learning rate. We then train the model for 10000 iterations.
 

docs/demos/pinn_forward/eulerbeam.rst

@@ -121,7 +121,7 @@ Next, we choose the network. Here, we use a fully connected neural network of de
     layer_size = [1] + [20] * 3 + [1]
     activation = "tanh"
     initializer = "Glorot uniform"
-    net = dde.maps.FNN(layer_size, activation, initializer)
+    net = dde.nn.FNN(layer_size, activation, initializer)
 
 Now, we have the PDE problem and the network. We build a ``Model`` and choose the optimizer and learning rate:
 

docs/demos/pinn_forward/helmholtz.2d.dirichlet.rst

@@ -131,7 +131,7 @@ Next, we choose the network. Here, we use a fully connected neural network of de
 
 .. code-block:: python
 
-  net = dde.maps.FNN(
+  net = dde.nn.FNN(
     [2] + [num_dense_nodes] * num_dense_layers + [1], activation, "Glorot uniform"
   )
 

docs/demos/pinn_forward/laplace.disk.rst

@@ -87,7 +87,7 @@ The number 2540 is the number of training residual points sampled inside the dom
 Next, we choose the network. Here, we use a fully connected neural network of depth 4 (i.e., 3 hidden layers) and width 20:
 
 .. code-block:: python
-    net = dde.maps.FNN([2] + [20] * 3 + [1], "tanh", "Glorot normal")
+    net = dde.nn.FNN([2] + [20] * 3 + [1], "tanh", "Glorot normal")
 
 If we rewrite this problem in cartesian coordinates, the variables are in the form of :math:`[r\sin(\theta), r\cos(\theta)]`. We use them as features to satisfy the certain underlying physical constraints, so that the network is automatically periodic along the :math:`\theta` coordinate and the period is :math:`2\pi`.
 

docs/demos/pinn_forward/lotka.volterra.rst

@@ -80,7 +80,7 @@ We have 3000 training residual points inside the domain and 2 points on the boun
     layer_size = [1] + [64] * 6 + [2]
     activation = "tanh"
     initializer = "Glorot normal"
-    net = dde.maps.FNN(layer_size, activation, initializer)
+    net = dde.nn.FNN(layer_size, activation, initializer)
 
 This is a neural network of depth 7 with 6 hidden layers of width 50. We use :math:`\tanh` as the activation function. Since we expect to have periodic behavior in the Lotka-Volterra equation, we add a feature layer with :math:`\sin(kt)`. This forces the prediction to be periodic and therefore more accurate.
 

docs/demos/pinn_forward/ode.system.rst

@@ -86,7 +86,7 @@ Next, we choose the network. Here, we use a fully connected neural network of de
     layer_size = [1] + [50] * 3 + [2]
     activation = "tanh"
     initializer = "Glorot uniform"
-    net = dde.maps.FNN(layer_size, activation, initializer)
+    net = dde.nn.FNN(layer_size, activation, initializer)
 
 Now, we have the ODE problem and the network. We bulid a ``Model`` and choose the optimizer and learning rate: