Value Error while feeding in Neural Network

Question

I have programmed a multi-layer neural network but I'm getting an error while feeding my dimension into it. I'm getting a Value Error.

Here is The Code:

import pandas as pd
import tensorflow as tf
import matplotlib.pyplot as plt
import numpy as np
from sklearn import datasets
from sklearn import metrics
from sklearn import model_selection
from sklearn import preprocessing


# In[207]:

df =pd.read_csv("train_data.csv")


# In[252]:

target = df["target"]
feat=df.drop(['target','connection_id'],axis=1)
target[189]


# In[209]:

len(feature.columns)



# In[210]:

logs_path="Server_attack"


# In[211]:

#Hyperparameters
batch_size=100
learning_rate=0.5
training_epochs=10


# In[244]:

X=tf.placeholder(tf.float32,[None,41])
Y_=tf.placeholder(tf.float32,[None,3])
lr=tf.placeholder(tf.float32)


# In[245]:

#5Layer Neural Network
L=200
M=100
N=60
O=30


# In[257]:

#Weights and Biases
W1=tf.Variable(tf.truncated_normal([41,L],stddev=0.1))
B1=tf.Variable(tf.ones([L]))
W2=tf.Variable(tf.truncated_normal([L,M],stddev=0.1))
B2=tf.Variable(tf.ones([M]))
W3=tf.Variable(tf.truncated_normal([M,N],stddev=0.1))
B3=tf.Variable(tf.ones([N]))
W4=tf.Variable(tf.truncated_normal([N,O],stddev=0.1))
B4=tf.Variable(tf.ones([O]))
W5=tf.Variable(tf.truncated_normal([O,3],stddev=0.1))
B5=tf.Variable(tf.ones([3]))               



# In[247]:

Y1=tf.nn.relu(tf.matmul(X,W1)+B1)
Y2=tf.nn.relu(tf.matmul(Y1,W2)+B2)
Y3=tf.nn.relu(tf.matmul(Y2,W3)+B3)
Y4=tf.nn.relu(tf.matmul(Y3,W4)+B4)
Ylogits=tf.nn.relu(tf.matmul(Y4,W5)+B5)
Y=tf.nn.softmax(Ylogits)


# In[216]:

cross_entropy = tf.nn.softmax_cross_entropy_with_logits(logits=Ylogits,labels=Y_)
cross_entropy = tf.reduce_mean(cross_entropy)


# In[217]:

correct_prediction=tf.equal(tf.argmax(Y,1),tf.argmax(Y_,1))
accuracy=tf.reduce_mean(tf.cast(correct_prediction,tf.float32))


# In[218]:

train_step=tf.train.AdamOptimizer(learning_rate).minimize(cross_entropy)


# In[219]:

#TensorBoard Parameters
tf.summary.scalar("cost",cross_entropy)
tf.summary.scalar("accuracy",accuracy)
summary_op=tf.summary.merge_all()


# In[220]:

init = tf.global_variables_initializer()
sess=tf.Session()
sess.run(init)


# In[253]:

with tf.Session() as sess:
    sess.run(init)
    writer = tf.summary.FileWriter(logs_path,graph=tf.get_default_graph())
    for epoch in range(training_epochs):
        batch_count=int(len(feature)/batch_size)
        for i in range(batch_count):


            batch_x,batch_y=feature.iloc[i, :].values.tolist(),target[i]

            _,summary = sess.run([train_step,summary_op],
                                 {X:batch_x,Y:batch_y,learning_rate:0.001}
                                )

I'm getting the following error:

ValueError: Cannot feed value of shape (41,) for Tensor 'Placeholder_24:0', which has shape '(?, 41)'

I need to reshape I guess.

score 0 · Answer 1 · answered Nov 01 '17 at 20:41

Your data format is incompatible with the placeholder defined as

X=tf.placeholder(tf.float32,[None,41])

It is probably easier to reformat your data which you feed during training/evaluation. I don't see where you import it but you are going to need to either reshape or swap axes so that it has format (index, 41) and not (41, index)

score 0 · Accepted Answer · answered Nov 01 '17 at 20:42

0

You're right, you just have to reshape your input values in order to make them compatible with the placeholder's shape.

Your placeholder has shape (?,41) that means any batch size, with 41 values. Your input is, instead, with a shape of 41.

It's clear that the batch dimension is missing. Just add a 1 dimension to your input and you'll be fine:

batch_x = np.expand_dims(np.array(feature.iloc[i, :].values.tolist()), axis=0)

Note that probably you have to add a 1 dimension to your batch_y variable too. (for the same reason described above)

answered Nov 01 '17 at 20:42

nessuno

26,493
5
83
74

Yes ,I 'm facing issue for batch_y now : ValueError: Cannot feed value of shape (1,) for Tensor 'Softmax_13:0', which has shape '(41, 3).I Tried to expand same way as you guided but resulted into error' – vidit02100 Nov 01 '17 at 20:50
I tried making similar change : batch_y = np.expand_dims(np.array(target[i]),axis=0) resulting to above error – vidit02100 Nov 01 '17 at 21:02
Your label is a single value, a scalar value in 0, 1, 2, isn't it? You need to one hot encode it first, in order to get a tensor with shape `3`. Then you have to extract a single value from the `target[i]` array to use the right (and single) label for the one same you feed into the network. However, you could ask a new question for this second problem. – nessuno Nov 01 '17 at 21:15
Sir I one hot encoded it But i don't know to retrieve it and get the tensor in shape. – vidit02100 Nov 01 '17 at 22:01
https://stackoverflow.com/questions/47068683/feeding-labels-with-one-hot-encoded-vectors-in-neural-network – vidit02100 Nov 02 '17 at 07:08

Value Error while feeding in Neural Network

2 Answers2

Linked