Convert a tensor to numpy array in Tensorflow?

TensorFlow 2.x

Eager Execution is enabled by default, so just call .numpy() on the Tensor object.

import tensorflow as tf

a = tf.constant([[1, 2], [3, 4]])                 
b = tf.add(a, 1)

a.numpy()
# array([[1, 2],
#        [3, 4]], dtype=int32)

b.numpy()
# array([[2, 3],
#        [4, 5]], dtype=int32)

tf.multiply(a, b).numpy()
# array([[ 2,  6],
#        [12, 20]], dtype=int32)

See NumPy Compatibility for more. It is worth noting (from the docs),

Numpy array may share memory with the Tensor object. Any changes to one may be reflected in the other.

Bold emphasis mine. A copy may or may not be returned, and this is an implementation detail based on whether the data is in CPU or GPU (in the latter case, a copy has to be made from GPU to host memory).

But why am I getting AttributeError: 'Tensor' object has no attribute 'numpy'?.
A lot of folks have commented about this issue, there are a couple of possible reasons:

  • TF 2.0 is not correctly installed (in which case, try re-installing), or
  • TF 2.0 is installed, but eager execution is disabled for some reason. In such cases, call tf.compat.v1.enable_eager_execution() to enable it, or see below.

If Eager Execution is disabled, you can build a graph and then run it through tf.compat.v1.Session:

a = tf.constant([[1, 2], [3, 4]])                 
b = tf.add(a, 1)
out = tf.multiply(a, b)

out.eval(session=tf.compat.v1.Session())    
# array([[ 2,  6],
#        [12, 20]], dtype=int32)

See also TF 2.0 Symbols Map for a mapping of the old API to the new one.

Related Posts:

  1. Convert a tensor to numpy array in Tensorflow?
  2. Convert a tensor to numpy array in Tensorflow?
  3. What is the difference between np.mean and tf.reduce_mean?
  4. RuntimeError: module compiled against API version 0xc but this version of numpy is 0xb
  5. ImportError: DLL load failed: The specified module could not be found
  6. ImportError: DLL load failed: The specified module could not be found
  7. ValueError: setting an array element with a sequence
  8. What exactly does numpy.exp() do? [closed]
  9. ImportError: No module named tensorflow
  10. How to find which version of TensorFlow is installed in my system?
  11. numpy: Invalid value encountered in true_divide
  12. ImportError: No module named ‘tensorflow.python’
  13. python numpy ValueError: operands could not be broadcast together with shapes
  14. ValueError: Unknown label type: ‘continuous’
  15. TypeError: unhashable type: ‘numpy.ndarray’
  16. How to fix IndexError: invalid index to scalar variable
  17. Could not install packages due to a “Environment error :[error 13]: permission denied : ‘usr/local/bin/f2py'”
  18. How does numpy.newaxis work and when to use it?
  19. Plotting a 2D heatmap with Matplotlib
  20. Should I use np.absolute or np.abs?
  21. What does numpy.random.seed(0) do?
  22. What does the c underscore expression `c_` do exactly?
  23. Convert pandas dataframe to NumPy array
  24. Unable to plot Double Bar, Bar plot using pyplot for ndarray
  25. Error: all the input array dimensions except for the concatenation axis must match exactly
  26. Singular matrix issue with Numpy
  27. How to find all occurrences of an element in a list
  28. TypeError: ‘numpy.float64’ object is not callable
  29. ‘DataFrame’ object has no attribute ‘sort’
  30. ‘DataFrame’ object has no attribute ‘sort’
  31. ValueError: all the input arrays must have same number of dimensions
  32. TypeError: cannot unpack non-iterable int objec
  33. ValueError: setting an array element with a sequence
  34. filename.whl is not a supported wheel on this platform
  35. How to normalize a NumPy array to a unit vector?
  36. How to raise a numpy array to a power? (corresponding to repeated matrix multiplications, not elementwise)
  37. What is dtype(‘O’), in pandas?
  38. How can I upgrade NumPy?
  39. What does numpy.gradient do?
  40. How can the Euclidean distance be calculated with NumPy?
  41. How to get element-wise matrix multiplication (Hadamard product) in numpy?
  42. Cannot uninstall Tensorflow
  43. ‘list’ object has no attribute ‘shape’
  44. What is dtype(‘O’), in pandas?
  45. ValueError: could not broadcast input array from shape (224,224,3) into shape (224,224)
  46. Overflow Error in Python’s numpy.exp function
  47. load csv into 2D matrix with numpy for plotting
  48. ValueError: ‘object too deep for desired array’
  49. Purpose of `numpy.log1p( )`?
  50. How to take column-slices of dataframe in pandas
  51. Pytorch tensor to numpy array
  52. ‘numpy.float64’ object is not iterable
  53. index 1 is out of bounds for axis 0 with size 1
  54. Normalize data in pandas
  55. Most efficient way to reverse a numpy array
  56. Can’t broadcast input array from shape (3,1) into shape (3,)
  57. Import error when trying to import tensorflow with gpu
  58. deleting rows in numpy array
  59. Failed to load the native TensorFlow runtime – TensorFlow 2.1
  60. How to normalize a 2-dimensional numpy array in python less verbose?
  61. Understanding NumPy’s einsum
  62. For loop and ‘numpy.float64’ object is not iterable error
  63. What are the causes of overflow encountered in double_scalars besides division by zero?
  64. How can I remove Nan from list Python/NumPy
  65. RuntimeWarning: numpy.dtype size changed, may indicate binary incompatibility
  66. TensorFlow: Blas GEMM launch failed
  67. TypeError: Invalid dimensions for image data when plotting array with imshow()
  68. What’s the difference between scikit-learn and tensorflow? Is it possible to use them together?
  69. tar: Unrecognized archive format error when trying to unpack flower_photos.tgz, TF tutorials on OSX
  70. Moving average or running mean
  71. can’t open tensorboard 0.0.0.0:6006 or localhost:6006
  72. How to start from second index for for-loop
  73. Using numpy to build an array of all combinations of two arrays
  74. Can’t update to numpy 1.13 with anaconda?
  75. How to plot an array in python?
  76. TypeError: only integer arrays with one element can be converted to an index 3
  77. How to create a numpy array of all True or all False?
  78. Removing nan values from an array
  79. LinAlgError: Last 2 dimensions of the array must be square
  80. Numpy, multiply array with scalar
  81. Building multi-regression model throws error: `Pandas data cast to numpy dtype of object. Check input data with np.asarray(data).`
  82. Python 3: Multiply a vector by a matrix without NumPy
  83. ImportError in importing from sklearn: cannot import name check_build
  84. How to add a new row to an empty numpy array
  85. Root mean square of a function in python
  86. What are the differences between numpy arrays and matrices? Which one should I use?
  87. mean, nanmean and warning: Mean of empty slice
  88. Is there any numpy group by function?
  89. Replacing Pandas or Numpy Nan with a None to use with MysqlDB
  90. Mean Squared Error in Numpy?
  91. How to normalize a NumPy array to within a certain range?
  92. Is there a head and tail method for Numpy array?
  93. How to implement the ReLU function in Numpy
  94. Sorting arrays in NumPy by column
  95. numpy : calculate the derivative of the softmax function
  96. Transposing a 1D NumPy array
  97. numpy.float64 object is not iterable…but I’m NOT trying to
  98. AttributeError: ‘module’ object has no attribute ‘main’ for tf.app.run()
  99. Conditional indexing with Numpy ndarray
  100. python numpy machine epsilon

Leave a Comment

tech