From 9de9c2627ea934b26ee675c23c17954e200ac26d Mon Sep 17 00:00:00 2001 From: MewX Date: Wed, 2 Aug 2017 14:20:47 +0930 Subject: [PATCH] Added data exception checker --- MachineLearning/TensorFlow/lltm_mlp.py | 16 +++++++++++++--- 1 file changed, 13 insertions(+), 3 deletions(-) diff --git a/MachineLearning/TensorFlow/lltm_mlp.py b/MachineLearning/TensorFlow/lltm_mlp.py index d4439f5..260db7d 100644 --- a/MachineLearning/TensorFlow/lltm_mlp.py +++ b/MachineLearning/TensorFlow/lltm_mlp.py @@ -9,7 +9,7 @@ import matplotlib.pyplot as plt # definitions -FILE_NAME = "newdata0725_variable_interval.csv" +FILE_NAME = "newdata.csv" # FILE_NAME = "test.csv" TIME_INTERVAL = 10 # records per second @@ -21,7 +21,7 @@ LEARNING_RATE = 0.003 STANDARD_DEVIATION = 0.1 -TRAINING_EPOCHS = 1000 +TRAINING_EPOCHS = 10 BATCH_SIZE = 50 # 100 DISPLAY_STEP = 20 RANDOM_STATE = 100 @@ -47,6 +47,12 @@ def find_second_beg(data_set, start_second): return start_idx +def outputSpecialData(arr, val): + for i in range(len(arr)): + if arr[i] > val or arr[i] < -val: + print("" + i + arr[i]) + + # fetch data from csv file # data format: # [0]: time @@ -62,7 +68,7 @@ def find_second_beg(data_set, start_second): raw_linear_heave = np.array(raw_data[:, 2]) raw_real_heave = np.array(raw_data[:, 3]) raw_nonlinear_heave = np.subtract(raw_real_heave, raw_linear_heave) -print(raw_nonlinear_heave[0]) +# outputSpecialData(raw_nonlinear_heave, 1) print(raw_data[find_second_beg(raw_data, 101)]) print(raw_data[find_second_beg(raw_data, 120)]) @@ -76,6 +82,7 @@ def find_second_beg(data_set, start_second): training_target[i] = [raw_nonlinear_heave[training_data_idx_start + i]] training_input = np.array(training_input) training_target = np.array(training_target) +# outputSpecialData(training_target, 1) print(training_input.shape) print(training_target.shape) @@ -87,9 +94,11 @@ def find_second_beg(data_set, start_second): testing_target[i] = [raw_nonlinear_heave[testing_data_idx_start + i]] testing_input = np.array(testing_input) testing_target = np.array(testing_target) +# outputSpecialData(testing_target, 1) print(testing_input.shape) print(testing_target.shape) + # NRMSE def nrmse(real, predict): up = tf.sqrt(tf.reduce_sum(tf.square(real - predict))) @@ -218,6 +227,7 @@ def mlp(_x, _weights, _biases): # Testing test_acc = sess.run(pred, feed_dict={X: testing_input, y: testing_target, dropout_keep_prob: 1.}) # print("Test accuracy: %.6f" % test_acc) +outputSpecialData(testing_target, 1) print(repr(np.column_stack((test_acc, testing_target)))) # for i in np.column_stack((test_acc, testing_target)): # print(repr(i))