>>> import tensorflow as tf 
>>> tens1 = tf.constant([[[1,2],[2,3]],[[3,4],[5,6]]]) 
>>> print sess.run(tens1)[1,1,0] 
5 

will throw an error

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
NameError: name 'sess' is not defined

Please create the sess object, before calling the run with either

sess = tf.Session()

or

sess = tf.InteractiveSession()

But these concepts are discussed only down below, so not sure if thats why it was intentionally ignored here.

I'm new to TensorFlow, hence can't figure out the error myself.

In Chapter-1 under Tensor Slicing and Joining section, I simply copied-pasted the code to see the results and I get the error.

Please help as my code is below:

import tensorflow as tf 
sess = tf.InteractiveSession()

t_matrix = tf.constant([[1,2,3],
                       [4,5,6],
                       [7,8,9]])
t_array = tf.constant([1,2,3,4,9,8,6,5])
t_array2 = tf.constant([2,3,4,5,6,7,8,9])

tf.split(0, 2, t_array) #splitting the array in two <--- Err in This Line

I get error while executing the split function as follows:

ValueError: Input must be scalar but has rank 1 for 'split_10' (op: 'Split') with input shapes: [8], [] and with computed input tensors: input[0] = <1 2 3 4 9 8 6 5>.

大佬，
这个部分的代码不能复现，错误太多了，
很多方法都变了，请更新下代码啊。
python3.6.6
tensorflow-1.12.0

section: Logistic function as a linear modeling generalization

actual:
The logistic function δ(t) is defined as follows

expected:
The logistic function σ(t) is defined as follows

There is no data of the example 2 in chapter 7

Hi,

I have couple of question on Chapter 7:

1. In Example 1 - univariate time series prediction with energy consumption data - when I run the example I get an error stating: AttributeError: module 'tensorflow.contrib.learn' has no attribute 'TensorFlowEstimator. Googling the problem I found tensflow have removed TensorFlowEstimator. What is the alternative for this?

2. No code supplied for the Example 2 on the github.

Great book, I've been following the examples. I'm new to ML so excuse my questions if they are silly.

1. In your quadratic example in chapter 5, the following line for dsY, I believe should not have the zero in front of it.
   testX, testY = dsX[trainsamples:trainsamples + testsamples], dsY[0:trainsamples:trainsamples+testsamples]

2. I noticed that you called one of your notebooks CH5 linear regression nn (neural network). However, in the actual chapter, it's called Non linear synthetic function regression. I'm sort of confused. Can you explain why you are calling it linear here and non-linear in the actual book?

section: Loss function description and Optimizer loop

now:
fori, j in zip(xvalues, yvalues)

expected:
for i, j in zip(xvalues, yvalues)

in Multivariate Linear Regression.py, it will not show the plot until I add one line:
plot.show()
Could you tell me why?

actual: hdf5
expected: hdfs

section: loss function

actual:
In the previous section we saw our approximated [latex]\hat p[/latex] function,

expected:
Is [latex]\hat p[/latex] right?

None of the link is working