３ヶ月コース＠鮎川

Lecture_TUKR/ayukawa_TUKR.py

@@ -0,0 +1,115 @@
+import numpy as np
+import jax,jaxlib
+import jax.numpy as jnp
+from tqdm import tqdm #プログレスバーを表示させてくれる
+
+
+class UKR:
+    def __init__(self, X, latent_dim, sigma, prior='random', Zinit=None):
+        #--------初期値を設定する．---------
+        self.X = X
+        #ここから下は書き換えてね
+        self.nb_samples, self.ob_dim = X.shape
+        self.sigma =sigma
+        self.latent_dim =latent_dim
+
+        if Zinit is None:
+            if prior == 'random': #一様事前分布のとき
+                self.Z = np.random.uniform(0, self.sigma*0.001, (self.nb_samples, self.latent_dim))
+                # Z1_vec = np.random.uniform(low=-1, high=1, size=Z)
+                # Z1_colum_vec = np.random.uniform(low=-1, high=1, size=[Z, 1])
+            # else: #ガウス事前分布のとき
+                # else: #Zの初期値が与えられた時
+            #self.Z = Zinit
+
+        self.history = {}
+
+    def kernel(self, Z1, Z2): #写像の計算 TUKRの式に変更
+            Mom = jnp.sum((Z1[:, None, :] - Z2[None, :, :]) ** 2, axis=2)
+            Chi = jnp.exp(-1/(2*self.sigma**2)*Mom)
+            f = (Chi@self.X)/jnp.sum(Chi, axis=1, keepdims=True)
+
+            return f
+
+    def E(self, Z, X, alpha, norm): #目的関数の計算
+        E = np.sum((X - self.kernel(Z,Z))**2)
+        R = alpha * jnp.sum(jnp.abs(Z ** norm))
+        E = E / self.nb_samples + R / self.nb_samples
+
+        return E
+
+    def fit(self, nb_epoch: int, eta: float, alpha: float, norm: float) :
+        # 学習過程記録用
+        self.history['z'] = np.zeros((nb_epoch, self.nb_samples, self.latent_dim))
+        self.history['kernel'] = np.zeros((nb_epoch, self.nb_samples, self.ob_dim))
+        self.history['error'] = np.zeros(nb_epoch)
+
+        for epoch in tqdm(range(nb_epoch)):
+            # Zの更新
+            dEdx = jax.grad(self.E, argnums=0)(self.Z, self.X, alpha, norm)
+            self.Z -= (eta) * dEdx
+
+
+            # 学習過程記録用
+            self.history['z'][epoch] = self.Z
+            self.history['kernel'][epoch] = self.kernel(self.Z,self.Z)
+            self.history['error'][epoch] = self.E(self.Z,self.X, alpha, norm)
+
+    #--------------以下描画用(上の部分が実装できたら実装してね)---------------------
+    def calc_approximate_f(self, resolution): #fのメッシュ描画用，resolution:一辺の代表点の数
+        nb_epoch = self.history['z'].shape[0]
+        self.history['y'] = np.zeros((nb_epoch, resolution ** self.latent_dim, self.ob_dim))
+        for epoch in tqdm(range(nb_epoch)):
+            zeta = create_zeta(self.Z, resolution)
+            Y = self.kernel(zeta, self.history['z'][epoch])
+            self.history['y'][epoch] = Y
+        return self.history['y']
+
+
+def create_zeta(Z, resolution): #fのメッシュの描画用に潜在空間に代表点zetaを作る．
+    z_x = np.linspace(np.min(Z), np.max(Z), resolution).reshape(-1, 1)
+    z_y = np.linspace(np.min(Z), np.max(Z), resolution)
+    XX, YY = np.meshgrid(z_x, z_y)
+    xx = XX.reshape(-1)
+    yy = YY.reshape(-1)
+    zeta = np.concatenate([xx[:, None], yy[:, None]], axis=1)
+
+
+
+
+
+    return zeta
+
+
+if __name__ == '__main__':
+    from data_scratch.data import load_kura_tsom
+    from Lecture_UKR.data import create_rasen
+    from Lecture_UKR.data import create_2d_sin_curve
+    from visualizer import visualize_history
+
+    #各種パラメータ変えて遊んでみてね．
+    ##
+    epoch = 300 #学習回数
+    sigma = 0.4 #カーネルの幅
+    eta = 2 #学習率
+    latent_dim = 2 #潜在空間の次元
+
+    alpha = 0
+    norm = 10
+
+    seed = 2
+    np.random.seed(seed)
+
+    #入力データ（詳しくはdata.pyを除いてみると良い）
+    nb_samples = 200 #データ数
+    X = load_kura_tsom(nb_samples) #鞍型データ　ob_dim=3, 真のL=2
+    # X = create_rasen(nb_samples) #らせん型データ　ob_dim=3, 真のL=1
+    # X = create_2d_sin_curve(nb_samples) #sin型データ　ob_dim=2, 真のL=1
+
+    ukr = UKR(X, latent_dim, sigma, prior='random')
+    ukr.fit(epoch, eta, alpha, norm)
+    #visualize_history(X, ukr.history['kernel'], ukr.history['z'], ukr.history['error'], save_gif=False, filename="tmp")
+    #----------描画部分が実装されたらコメントアウト外す----------
+    ukr.calc_approximate_f(resolution=10)
+    visualize_history(X, ukr.history['y'], ukr.history['z'], ukr.history['error'], save_gif=False, filename="tmp")
+

Lecture_TUKR/data_scratch.py

@@ -2,18 +2,18 @@
 import matplotlib.pyplot as plt
 
 def load_kura_tsom(xsamples, ysamples, missing_rate=None,retz=False):
-    z1 = 
-    z2 = 
+    z1 = np.linspace(-1, 1, xsamples)
+    z2 = np.linspace(-1, 1, ysamples)
 
-    z1_repeated, z2_repeated = 
-    x1 = 
-    x2 = 
-    x3 = 
+    z1_repeated, z2_repeated = np.meshgrid(z1, z2)
+    x1 = z1_repeated
+    x2 = z2_repeated
+    x3 = x1**2 - x2**2
     #ノイズを加えたい時はここをいじる,locがガウス分布の平均、scaleが分散,size何個ノイズを作るか
     #このノイズを加えることによって三次元空間のデータ点は上下に動く
 
-    x = 
-    truez = 
+    x = np.concatenate((x1[:, :, np.newaxis], x2[:, :, np.newaxis], x3[:, :, np.newaxis]), axis=2)
+    truez = np.concatenate((z2_repeated[:, :, np.newaxis], z2_repeated[:, :, np.newaxis]), axis=2)
 
     if missing_rate == 0 or missing_rate == None:
         if retz:
@@ -25,13 +25,13 @@ def load_kura_tsom(xsamples, ysamples, missing_rate=None,retz=False):
     import matplotlib.pyplot as plt
     from mpl_toolkits.mplot3d import Axes3D
 
-    xsamples = 
-    ysamples = 
+    xsamples = 20
+    ysamples = 20
 
-    x, truez = load_kura_tsom() 
+    x, truez = load_kura_tsom(xsamples, ysamples, retz=True)
 
     fig = plt.figure(figsize=[5, 5])
     ax_x = fig.add_subplot(projection='3d')
-    ax_x.scatter()
+    ax_x.scatter(x[:, :, 0], x[:, :, 1], x[:, :, 2])
     ax_x.set_title('Generated three-dimensional data')
     plt.show()

Lecture_TUKR/sample/tukr_kura_data.py

@@ -14,6 +14,7 @@ def load_kura_tsom(xsamples, ysamples, missing_rate=None,retz=False):
     #ノイズを加えたい時はここをいじる,locがガウス分布の平均、scaleが分散,size何個ノイズを作るか
     #このノイズを加えることによって三次元空間のデータ点は上下に動く
 
+
     x = np.concatenate((x1[:, :, np.newaxis], x2[:, :, np.newaxis], x3[:, :, np.newaxis]), axis=2)
     truez = np.concatenate((z1_repeated[:, :, np.newaxis], z2_repeated[:, :, np.newaxis]), axis=2)
 

Lecture_TUKR/test.py

@@ -1 +1,2 @@
-import numpy as np
+import numpy as np
+print("hello")