Merge pull request #34 from clinfo/loss_plot

Loss plot

Author: kojima-r

script/plot_loss.py

@@ -0,0 +1,58 @@
+import matplotlib.pyplot as plt
+import sys
+
+def plot_loss(loss_txt,loss_image):
+    f = open(loss_txt,"r")
+    datalist=f.readlines()
+
+    epoch = []
+    train_cost = []
+    train_cost_recons = []
+    train_cost_temp = []
+    train_cost_pred = []
+    valid_cost = []
+    valid_cost_recons = []
+    valid_cost_temp = []
+    valid_cost_pred = []
+
+    for data in datalist[1:]:
+        a = data.split('\t')
+        epoch.append(a[1])
+        train_cost.append(float(a[2]))
+        train_cost_recons.append(float(a[9]))
+        train_cost_temp.append(float(a[10])*0.5)
+        train_cost_pred.append(float(a[11])*0.1)
+        valid_cost.append(float(a[3]))
+        valid_cost_recons.append(float(a[12]))
+        valid_cost_temp.append(float(a[13])*0.5)
+        valid_cost_pred.append(float(a[14])*0.1)
+
+    plt.style.use("grayscale")
+    fig = plt.figure(figsize=(16.0,6.0))
+    ax1 = fig.add_subplot(1, 2, 1)
+    ax1.set_title("train cost")
+    ax1.set_xlabel("epoch")
+    ax1.set_xlim([0,1000])
+    ax1.set_ylim([0.01,5.0])
+    ax1.plot(train_cost,label="total")
+    ax1.plot(train_cost_recons,label="recons")
+    ax1.plot(train_cost_temp,label="alpha*temporal")
+    ax1.plot(train_cost_pred,label="beta*prediction")
+    ax1.legend()
+
+    ax2 = fig.add_subplot(1, 2, 2)
+    ax2.set_title("valid cost")
+    ax2.set_xlabel("epoch")
+    ax2.set_xlim([0,1000])
+    ax2.set_ylim([0.01,5.0])
+    ax2.plot(train_cost,label="total")
+    ax2.plot(train_cost_recons,label="recons")
+    ax2.plot(train_cost_temp,label="alpha*temporal")
+    ax2.plot(train_cost_pred,label="beta*prediction")
+    ax2.legend()
+
+    plt.savefig(loss_image)
+
+if __name__ == '__main__':
+    args = sys.argv
+    plot_loss(args[1],args[2])