From f0369909d0c5cc5c9cccbcf98b09c16fadea5457 Mon Sep 17 00:00:00 2001
From: Junwei Pan <pandevirus@gmail.com>
Date: Sun, 8 Jan 2017 15:34:06 -0800
Subject: [PATCH] Style Fix (#4923)

* Style Fix

* Style Fix
---
 docs/autogen.py                      | 2 +-
 examples/deep_dream.py               | 2 +-
 examples/neural_doodle.py            | 2 +-
 examples/neural_style_transfer.py    | 2 +-
 keras/layers/advanced_activations.py | 6 +++---
 keras/utils/layer_utils.py           | 2 +-
 6 files changed, 8 insertions(+), 8 deletions(-)

diff --git a/docs/autogen.py b/docs/autogen.py
index 826526aff..679dbc193 100644
--- a/docs/autogen.py
+++ b/docs/autogen.py
@@ -312,7 +312,7 @@ def get_function_signature(function, method=True):
     for a in args:
         st += str(a) + ', '
     for a, v in kwargs:
-        if type(v) == str:
+        if isinstance(v, str):
             v = '\'' + v + '\''
         st += str(a) + '=' + str(v) + ', '
     if kwargs or args:
diff --git a/examples/deep_dream.py b/examples/deep_dream.py
index 2e3f7c08f..8826e1d24 100644
--- a/examples/deep_dream.py
+++ b/examples/deep_dream.py
@@ -140,7 +140,7 @@ loss += settings['dream_l2'] * K.sum(K.square(dream)) / np.prod(img_size)
 grads = K.gradients(loss, dream)
 
 outputs = [loss]
-if type(grads) in {list, tuple}:
+if isinstance(grads, (list, tuple)):
     outputs += grads
 else:
     outputs.append(grads)
diff --git a/examples/neural_doodle.py b/examples/neural_doodle.py
index 890b208ee..40536544c 100644
--- a/examples/neural_doodle.py
+++ b/examples/neural_doodle.py
@@ -301,7 +301,7 @@ loss_grads = K.gradients(loss, target_image)
 
 # Evaluator class for computing efficiency
 outputs = [loss]
-if type(loss_grads) in {list, tuple}:
+if isinstance(loss_grads, (list, tuple)):
     outputs += loss_grads
 else:
     outputs.append(loss_grads)
diff --git a/examples/neural_style_transfer.py b/examples/neural_style_transfer.py
index 8f6b59ee0..53397dce1 100644
--- a/examples/neural_style_transfer.py
+++ b/examples/neural_style_transfer.py
@@ -208,7 +208,7 @@ loss += total_variation_weight * total_variation_loss(combination_image)
 grads = K.gradients(loss, combination_image)
 
 outputs = [loss]
-if type(grads) in {list, tuple}:
+if isinstance(grads, (list, tuple)):
     outputs += grads
 else:
     outputs.append(grads)
diff --git a/keras/layers/advanced_activations.py b/keras/layers/advanced_activations.py
index 2caf3059e..cf85302cd 100644
--- a/keras/layers/advanced_activations.py
+++ b/keras/layers/advanced_activations.py
@@ -71,7 +71,7 @@ class PReLU(Layer):
         self.supports_masking = True
         self.init = initializations.get(init)
         self.initial_weights = weights
-        if type(shared_axes) is not list and type(shared_axes) is not tuple:
+        if not isinstance(shared_axes, (list, tuple)):
             self.shared_axes = [shared_axes]
         else:
             self.shared_axes = list(shared_axes)
@@ -174,7 +174,7 @@ class ParametricSoftplus(Layer):
         self.alpha_init = K.cast_to_floatx(alpha_init)
         self.beta_init = K.cast_to_floatx(beta_init)
         self.initial_weights = weights
-        if type(shared_axes) is not list and type(shared_axes) is not tuple:
+        if not isinstance(shared_axes, (list, tuple)):
             self.shared_axes = [shared_axes]
         else:
             self.shared_axes = list(shared_axes)
@@ -279,7 +279,7 @@ class SReLU(Layer):
         self.a_left_init = a_left_init
         self.t_right_init = t_right_init
         self.a_right_init = a_right_init
-        if type(shared_axes) is not list and type(shared_axes) is not tuple:
+        if not isinstance(shared_axes, (list, tuple)):
             self.shared_axes = [shared_axes]
         else:
             self.shared_axes = list(shared_axes)
diff --git a/keras/utils/layer_utils.py b/keras/utils/layer_utils.py
index 00c0d49be..7b142323e 100644
--- a/keras/utils/layer_utils.py
+++ b/keras/utils/layer_utils.py
@@ -126,7 +126,7 @@ def count_total_params(layers, layer_set=None):
         if layer in layer_set:
             continue
         layer_set.add(layer)
-        if type(layer) in (Model, Sequential):
+        if isinstance(layer, (Model, Sequential)):
             t, nt = count_total_params(layer.layers, layer_set)
             trainable_count += t
             non_trainable_count += nt