Merge remote-tracking branch 'refs/remotes/origin/master' into Book

Author: carefree0910

a_FirstExample/README.md

@@ -4,4 +4,4 @@ Simple regression under **LSE** criterion using `np.polyfit` & `np.polyval`
 Dataset comes from Coursera's Machine Learning Course
 
 ## Result
-![image](http://i2.muimg.com/567571/f0cac9ec46fd2837.png)
+![image](http://oph1aen4o.bkt.clouddn.com/18-1-30/89962212.jpg)

b_NaiveBayes/README.md

@@ -8,9 +8,9 @@ Naive Bayes algorithms implemented with
 ## Visualization
 + Dataset comes from UCI: [Bank Marketing dataset](http://archive.ics.uci.edu/ml/datasets/Bank+Marketing)
 
-![Discrete Features](http://i4.buimg.com/567571/79ea5cb079b16a72.png)
+![Categorical Features](http://oph1aen4o.bkt.clouddn.com/18-1-30/53941771.jpg)
 
-![Continuous Features](http://i2.muimg.com/567571/f267b1a607e4c95e.png)
+![Numerical Features](http://oph1aen4o.bkt.clouddn.com/18-1-30/53966960.jpg)
 
 ## Example
 ```python
@@ -24,4 +24,4 @@ nb.evaluate(x, y)                       # Print out accuracy
 nb.visualize2d(x, y)                    # Visualize result (2d)
 nb.visualize()                          # Visualize distribution
 nb.show_timing_log()                    # Show timing log
-```
+```

c_CvDTree/README.md

@@ -5,13 +5,13 @@ Implemented **ID3**, **C4.5** & **CART**. Capable for dealing with relatively 'r
 + Dataset comes from UCI: [Mushroom dataset](http://archive.ics.uci.edu/ml/datasets/Mushroom)
 
 ### ID3
-![ID3](http://i1.piimg.com/567571/b202b2dfd1394757.png)
+![ID3](http://oph1aen4o.bkt.clouddn.com/18-1-30/99814955.jpg)
 
 ### C4.5
-![C4.5](http://i1.piimg.com/567571/d64bffa200033d00.png)
+![C4.5](http://oph1aen4o.bkt.clouddn.com/18-1-30/48437221.jpg)
 
 ### CART
-![CART](http://i1.piimg.com/567571/330a93ad355c0a05.png)
+![CART](http://oph1aen4o.bkt.clouddn.com/18-1-30/32419745.jpg)
 
 ## Example
 ```python
@@ -26,4 +26,4 @@ tree.evaluate(x, y)                     # Print out accuracy
 tree.visualize2d(x, y)                  # Visualize result (2d)
 tree.visualize()                        # Visualize Cart Tree itself
 tree.show_timing_log()                  # Show timing log
-```
+```

c_CvDTree/TestTree.py

@@ -110,5 +110,6 @@ def main(visualize=True):
 
     tree.show_timing_log()
 
+
 if __name__ == '__main__':
     main(False)

c_CvDTree/Tree.py

@@ -188,20 +188,24 @@ def view(self):
         self.root.view()
 
     @CvDBaseTiming.timeit(level=2, prefix="[API] ")
-    def visualize(self, radius=24, width=1200, height=800, padding=0.2, plot_num=30, title="CvDTree"):
+    def visualize(self, radius=24, width=1200, height=800,
+                  height_padding_ratio=0.2, width_padding=30, title="CvDTree"):
         self._update_layers()
-        units = [len(layer) for layer in self.layers]
+        n_units = [len(layer) for layer in self.layers]
 
         img = np.ones((height, width, 3), np.uint8) * 255
-        axis0_padding = int(height / (len(self.layers) - 1 + 2 * padding)) * padding + plot_num
-        axis0 = np.linspace(
-            axis0_padding, height - axis0_padding, len(self.layers), dtype=np.int)
-        axis1_padding = plot_num
-        axis1 = [np.linspace(axis1_padding, width - axis1_padding, unit + 2, dtype=np.int)
-                 for unit in units]
-        axis1 = [axis[1:-1] for axis in axis1]
-
-        for i, (y, xs) in enumerate(zip(axis0, axis1)):
+        height_padding = int(
+            height / (len(self.layers) - 1 + 2 * height_padding_ratio)
+        ) * height_padding_ratio + width_padding
+        height_axis = np.linspace(
+            height_padding, height - height_padding, len(self.layers), dtype=np.int)
+        width_axis = [
+            np.linspace(width_padding, width - width_padding, unit + 2, dtype=np.int)
+            for unit in n_units
+        ]
+        width_axis = [axis[1:-1] for axis in width_axis]
+
+        for i, (y, xs) in enumerate(zip(height_axis, width_axis)):
             for j, x in enumerate(xs):
                 if i == 0:
                     cv2.circle(img, (x, y), radius, (225, 100, 125), 1)
@@ -216,13 +220,13 @@ def visualize(self, radius=24, width=1200, height=800, padding=0.2, plot_num=30,
                     color = (0, 255, 0)
                 cv2.putText(img, text, (x-7*len(text)+2, y+3), cv2.LINE_AA, 0.6, color, 1)
 
-        for i, y in enumerate(axis0):
-            if i == len(axis0) - 1:
+        for i, y in enumerate(height_axis):
+            if i == len(height_axis) - 1:
                 break
-            for j, x in enumerate(axis1[i]):
-                new_y = axis0[i + 1]
+            for j, x in enumerate(width_axis[i]):
+                new_y = height_axis[i + 1]
                 dy = new_y - y - 2 * radius
-                for k, new_x in enumerate(axis1[i + 1]):
+                for k, new_x in enumerate(width_axis[i + 1]):
                     dx = new_x - x
                     length = np.sqrt(dx**2+dy**2)
                     ratio = 0.5 - min(0.4, 1.2 * 24/length)

d_Ensemble/README.md

@@ -16,4 +16,4 @@ rf.show_timing_log()                    # Show timing log
 ```
 
 ### Result
-![Cart on Spiral](http://i4.buimg.com/567571/fd50b722d988b1dd.png)
+![Cart on Spiral](http://oph1aen4o.bkt.clouddn.com/18-1-30/25490957.jpg)

e_SVM/KP.py

@@ -58,6 +58,7 @@ def _get_grads(self, x_batch, y_batch, y_pred, sample_weight_batch, *args):
             ]
         return np.sum(err[mask])
 
+
 if __name__ == '__main__':
     # xs, ys = DataUtil.gen_two_clusters(center=5, dis=1, scale=2, one_hot=False)
     xs, ys = DataUtil.gen_spiral(20, 4, 2, 2, one_hot=False)

e_SVM/Perceptron.py

@@ -60,3 +60,39 @@ def predict(self, x, get_raw_results=False, **kwargs):
         if get_raw_results:
             return rs
         return np.sign(rs).astype(np.float32)
+
+
+class Perceptron2(Perceptron):
+    def fit(self, x, y, sample_weight=None, lr=None, epoch=None, animation_params=None):
+        if sample_weight is None:
+            sample_weight = self._params["sample_weight"]
+        if lr is None:
+            lr = self._params["lr"]
+        if epoch is None:
+            epoch = self._params["epoch"]
+        *animation_properties, animation_params = self._get_animation_params(animation_params)
+
+        x, y = np.atleast_2d(x), np.asarray(y)
+        if sample_weight is None:
+            sample_weight = np.ones(len(y))
+        else:
+            sample_weight = np.asarray(sample_weight) * len(y)
+
+        self._w = np.random.random(x.shape[1])
+        self._b = 0.
+        ims = []
+        bar = ProgressBar(max_value=epoch, name="Perceptron")
+        for i in range(epoch):
+            y_pred = self.predict(x, True)
+            err = -y * y_pred * sample_weight
+            idx = np.argmax(err)
+            if err[idx] < 0:
+                bar.terminate()
+                break
+            w_norm = np.linalg.norm(self._w)
+            delta = lr * y[idx] * sample_weight[idx] / w_norm
+            self._w += delta * (x[idx] - y_pred[idx] * self._w / w_norm ** 2)
+            self._b += delta
+            self._handle_animation(i, x, y, ims, animation_params, *animation_properties)
+            bar.update()
+        self._handle_mp4(ims, animation_properties)

e_SVM/TestLinear.py

@@ -43,5 +43,6 @@ def main():
 
     perceptron.show_timing_log()
 
+
 if __name__ == '__main__':
     main()

e_SVM/TestSVM.py

@@ -110,5 +110,6 @@ def main():
 
     svm.show_timing_log()
 
+
 if __name__ == '__main__':
     main()