Why do I get the message “Python DeprecationWrapper with no properties” when I try to load a Python module in MATLAB?

Question

MathWorks Support Team 2021-4-20

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/807972-why-do-i-get-the-message-python-deprecationwrapper-with-no-properties-when-i-try-to-load-a-python

回答： MathWorks Support Team 2021-4-20

As an example, when I try to import TensorFlow (version 1.14.0) in MATLAB R2020b with Python 3.7.9 I get the following message

>> py.importlib.import_module('tensorflow')
ans = 
  Python DeprecationWrapper with no properties.
    <module 'tensorflow' from 'C:\\Users\\<username>\\AppData\\Local\\Programs\\Python\\Python37\\lib\\site-packages\\tensorflow\\__init__.py'>

Why am I getting this message and how can I load my module?

请先登录，再回答此问题。

Answer 1

MathWorks Support Team 2021-4-20

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/807972-why-do-i-get-the-message-python-deprecationwrapper-with-no-properties-when-i-try-to-load-a-python#answer_680142

This issue stems from a Python warning(s) that some features will be deprecated in the future. If you attempt to import a deprecated version of TensorFlow directly into Python you will see these types of messages:

Python 3.7.9 (tags/v3.7.9:13c94747c7, Aug 17 2020, 18:58:18) [MSC v.1900 64 bit (AMD64)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> import tensorflow
C:\Users\<username>\AppData\Local\Programs\Python\Python37\lib\site-packages\tensorflow\python\framework\dtypes.py:516: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_qint8 = np.dtype([("qint8", np.int8, 1)])
...

Attempting to import a module with deprecation warnings into MATLAB will return a Python DeprecationWrapper object with no properties. In these cases, the best solution is to update to a newer version of the module.

If updating to a newer version is not feasible, you can try the following workaround using Python’s getattr function. The example, below, uses the previous TensorFlow example with a workaround to access the TensorFlow constant function.

>> m = py.importlib.import_module('tensorflow');
>> f = py.getattr(m,'constant')
f = 
  Python function with no properties.
    <function constant_v1 at 0x0000026AA97FA798>
>> c = f(1)
c = 
  Python Tensor with properties:
         device: [1×0 py.str]
          dtype: [1×1 py.tensorflow.python.framework.dtypes.DType]
          graph: [1×1 py.tensorflow.python.framework.ops.Graph]
           name: [1×7 py.str]
             op: [1×1 py.tensorflow.python.framework.ops.Operation]
          shape: [1×1 py.tensorflow.python.framework.tensor_shape.TensorShape]
    value_index: [1×1 py.int]
    Tensor("Const:0", shape=(), dtype=float32)