Question

我想在TFLearn做张量流示例“波士顿住房价格”。但我得到了形状错误。

这是我的代码：

    import tflearn   
    from tflearn.data_utils import load_csv

    data, target = load_csv('boston_train.csv', has_header=True)  
    input_ = tflearn.input_data(shape=[None, 9])   
    linear = tflearn.fully_connected(input_, 9)   
    regression = tflearn.regression(linear, optimizer='sgd', loss='mean_square', learning_rate=0.01)     
    m = tflearn.DNN(regression)
    m.fit(data, target, n_epoch=10, batch_size=10, show_metric=True)

我收到以下错误：

    ValueError: Cannot feed value of shape (10,) for Tensor 'TargetsData/Y:0', which has shape '(?, 9)'

csv文件有9个功能和一个标签列我该怎么办？

Answer 1

感谢您的回答！

我解决了问题，所以这里是代码：

import numpy as np
import tflearn
from tflearn.data_utils import load_csv
from numpy import genfromtxt


data, target = load_csv('boston_train.csv', has_header=True)
target = np.reshape(target, (-1,1))

net = tflearn.input_data(shape=[None, 9])
net = tflearn.fully_connected(net, 9)
net = tflearn.fully_connected(net, 1)
net = tflearn.regression(net, optimizer='sgd', loss='mean_square', learning_rate=0.01)
net = tflearn.DNN(net)
net.fit(data, target, n_epoch=10, batch_size=10, show_metric=True)

test_data = genfromtxt('boston_predict.csv', delimiter=',', skip_header = 1)
test_data = np.reshape(test_data, (-1,9))

pred = net.predict(test_data)
print(pred)

Tensorflor / TFLearn不能提供形状的价值

1 个答案: