载入文本数据：如txt、csv格式等。 直接上码 例如： CSV数据如下，
import numpy as np import pandas as pd import tensorflow as tf fileData = pd.read_csv('D:\pythonxuni\data1.csv',dtype= np.float32, header= None) #将fileData设置为二维数组 wholeData = fileData.as_matrix() #wholeData获取的是整个数组中所有项的个数，共16项，wholeData[0].size为数组的第一行的项数，为4. rowCount = int(wholeData.size / wholeData[0].size) goodCount = 0 for i in range(rowCount): if wholeData[i][0] * 0.6 + wholeData[i][1] * 0.3 + wholeData[i][2] *0.1 >=95: goodCount += 1 print('wholeData = %s' % wholeData) print('rowCount = %d' % rowCount) print('goodCount = %d' % goodCount) x = tf.placeholder(shape= [3], dtype= tf.float32) yTrain = tf.placeholder(shape =[], dtype= tf.float32) w = tf.Variable(tf.zeros([3]), dtype= tf.float32) b = tf.Variable(80, dtype= tf.float32) #softmax函数将w的值相加为1 wn = tf.nn.softmax(w) n1 = x * wn n2 = tf.reduce_sum(n1) - b y = tf.nn.sigmoid(n2) loss = tf.abs(y - yTrain) optimizer = tf.train.RMSPropOptimizer(0.1) train = optimizer.minimize(loss) sess = tf.Session() init = tf.global_variables_initializer() sess.run(init) for i in range(2): for j in range(rowCount): result = sess.run([train, x, yTrain, wn, b, n2, y, loss], feed_dict={x:wholeData[j][0:3], yTrain: wholeData[j][3]}) print(result)