Keras: AttributeError: module 'keras.backend' has no attribute 'image_dim_ordering'

[mazatov]

After a recent update Keras backend cannot find image_dim_ordering parameter.

Has this been moved in the recent update?

import keras
c:\users\popeye\appdata\local\programs\python\python35\lib\site-packages\h5py\__init__.py:36: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.
  from ._conv import register_converters as _register_converters
Using TensorFlow backend.

keras.backend.image_dim_ordering()
Traceback (most recent call last):

  File "<ipython-input-2-78c718c47677>", line 1, in <module>
    keras.backend.image_dim_ordering()

AttributeError: module 'keras.backend' has no attribute 'image_dim_ordering'

[ymodak]

What keras and TensorFlow versions are you using?
I checked with Keras 2.2.4 and TF 1.13.1 and was able to execute
keras.backend.image_dim_ordering() successfully. Can you please confirm? Thanks!

[mazatov]

I recently updated to the bleeding edge version. It's only there that I have the issue.

[mazatov]

Also It seems to work with
K.common.image_dim_ordering()

[ymodak]

Did you try import keras before using the image ordering functions.

[ymodak]

Automatically closing due to lack of recent activity. Please update the issue when new information becomes available, and we will reopen the issue. Thanks!

[Genarito]

In Keras v2 the method has be renamed to image_data_format. So you coud try:

keras.backend.image_data_format()

Source

[aclex]

From my little version-wise research, this attribute disappears in 2.2.4 -> 2.2.5 version change. So in my case everything still works fine with 2.2.4.

Thanks everyone for suggestions to fix it!

[KhawlaSeddiki]

I am having the same error (AttributeError: module 'keras.backend' has no attribute 'image_dim_ordering') with version 2.3.1. Why?

[AayushShah25]

now use k.set_image_data_format('channels_first')
The latest version has replaced the image_dim_ordering to image_data_format.

And the function set_image_data_format() does the same as image_dim_ordering was doing.

( use 'channels_first' instead of 'tf' )

[breakcards]

Hello,
I am using following versions:
keras : 2.3.1
tf : 1.14.0

I have tried all the above solutions given, still not able to deal with the following error

`45
46 def __init__(self, gamma_init=20, **kwargs):
---> 47 if (K.set_image_data_format('channels_first')== 'tf'):
48 self.axis = 3
49 else:

AttributeError: module 'keras.backend' has no attribute 'image_dim_ordering'`

---

[TheHugeManatee]

@nikhilshinde095 I think in your case you need
if K.image_data_format() == 'channels_first':
because you are checking, not setting the image_data_format

[mserouar]

@TheHugeManatee And what about the old K.set_image_dim_ordering('th'), if we only have channel_first/last ?

[TheHugeManatee]

The way I understand it, with the new API:
setting:
K.set_image_dim_ordering('tf') --> K.set_image_data_format('channels_last')
K.set_image_dim_ordering('th') --> K.set_image_data_format('channels_first')

checking:
K.image_dim_ordering() == 'tf' --> K.image_data_format() == 'channels_last'
K.image_dim_ordering() == 'th' --> K.image_data_format() == 'channels_first'

I'm not very familiar with it so there might be other edge cases.

[Edited: Corrected the order as per @CarMiranda 's comment below]

[CarMiranda]

The way I understand it, with the new API:
setting:
K.set_image_dim_ordering('tf') --> K.set_image_data_format('channels_first')
K.set_image_dim_ordering('th') --> K.set_image_data_format('channels_last')

checking:
K.image_dim_ordering() == 'tf' --> K.image_data_format() == 'channels_first'
K.image_dim_ordering() == 'th' --> K.image_data_format() == 'channels_last'

I'm not very familiar with it so there might be other edge cases.

this should read:
K.set_image_dim_ordering('tf') --> K.set_image_data_format('channels_last')
K.set_image_dim_ordering('th') --> K.set_image_data_format('channels_first')
K.image_dim_ordering() == 'tf' --> K.image_data_format() == 'channels_last'
K.image_dim_ordering() == 'th' --> K.image_data_format() == 'channels_first'
(the other way around)

[TheHugeManatee]

You are right! tf, i.e. TensorFlow ordering, corresponds to channels_last.
I edited my original post to avoid confusion for any readers just cursory scrolling through..

[gavirapp]

I kept running into this problem for several months. Given I am no hack like many others here, I finally figured out it is because of inconsistent version of tensorflow and keras that I have been installing over the years and the major overhaul of tensorflow to ver 2.

1. Complete uninstall of the anaconda: even removing all the packages from .conda etc. (I kept a list of what I currently have using $conda list > installed_packages_4_21_20.txt)
2. Reinstalled anaconda with latest stable release
3. Started installing packages as needed (such as tensorflow, keras etc.) from Anaconda GUI with Environment interface
4. Some packages like opencv-python I had to install from CLI with pip command

No more "AttributeError: module 'keras.backend' has no attribute 'xxxxxxxxx' " for couple days now. Hope this helps. Just a warning, make sure you can afford to uninstall anaconda completely!

[kaavyaastalin]

Replace K.image_dim_ordering() to K.common.image_dim_ordering()

[Deeptanshu-sankhwar]

@kaavyaastalin thanks, this worked for me

[nantha42]

I get the following error in colab if I use K.common.image_dim_ordering(),
module 'tensorflow._api.v2.config' has no attribute 'experimental_list_devices'

[Harsha-1729]

The way I understand it, with the new API:
setting:
K.set_image_dim_ordering('tf') --> K.set_image_data_format('channels_last')
K.set_image_dim_ordering('th') --> K.set_image_data_format('channels_first')

checking:
K.image_dim_ordering() == 'tf' --> K.image_data_format() == 'channels_last'
K.image_dim_ordering() == 'th' --> K.image_data_format() == 'channels_first'

I'm not very familiar with it so there might be other edge cases.

[Edited: Corrected the order as per @CarMiranda 's comment below]

Thanks, It worked for me
Versions:
Keras - 2.3.1
TensorFlow - 1.14

[yashkpamnani]

What keras and TensorFlow versions are you using?
I checked with Keras 2.2.4 and TF 1.13.1 and was able to execute
keras.backend.image_dim_ordering() successfully. Can you please confirm? Thanks!

I'm using 2.3.1, and it's not working,

[Harsha-1729]

What keras and TensorFlow versions are you using?
I checked with Keras 2.2.4 and TF 1.13.1 and was able to execute
keras.backend.image_dim_ordering() successfully. Can you please confirm? Thanks!

I'm using 2.3.1, and it's not working,

Versions:
Keras - 2.3.1
TensorFlow - 1.14

[Salmanvaliyakath]

AttributeError: module 'keras.backend' has no attribute 'image_dim_ordering'

in keras 2.3.1

tried solution:-
1.K.image_dim_ordering() == 'tf' --> K.image_data_format() == 'channels_last' not working.

[Rakesh-MS-91]

Hey @Salmanvaliyakath

If you don't want to change your current environment. Then do the following changes in losses.py file, resnet.py file and in train_frcnn.py file

In Keras v2 the 'image_dim_ordering' method has been renamed to image_data_format.

So you could try:

K.image_data_format()` instead of K.image_dim_ordering()

if K.image_data_format() == 'th':

This solved my problem.