Cannot upgrade Python with Conda Package Manager on Ubuntu 22.04

[bredsj]

I have installed oneAPI with AI toolkits and work mainly in Fortran combined with Python. In the Conda base environment I could update Python to version=5.4.3 but when I select the intelenv it shows that an update is available but nothings happen when I update.

I have done the following from ~/anaconda3/bin

conda init

anaconda-navigator

and then in the navigator selected the intelenv, etc

When I try the terminal the following happens

steph@+++++:$ conda update anaconda

PackageNotInstalledError: Package is not installed in prefix.
prefix: /opt/intel/oneapi/intelpython/latest
package name: anaconda

steph@+++++:$

Any assistance would be appreciated.

[DiyaN_Intel]

Hi.

Thank you for posting in Intel Communities.

We are working on this internally and get back to you with an update soon.

Regards,

Diya

[Vasudha_Intel]

Hello @bredsj, could you please try conda update --all and let us know if it works for you?

[bredsj]

Please see result below, to confirm; I use Ubuntu 22.03 LTS, the first batch of feedback is what is loaded per session, then the command and the feedback. Because Intel is by default installed in /opt administrative rights (i.e. SUDO) is required. When SUDO is used the local rights are overridden, or at least it is my understanding.

thanks for 

:: initializing oneAPI environment ...
bash: BASH_VERSION = 5.1.16(1)-release
args: Using "$@" for setvars.sh arguments:
:: advisor -- latest
:: ccl -- latest
:: clck -- latest
:: compiler -- latest
:: dal -- latest
:: debugger -- latest
:: dev-utilities -- latest
:: dnnl -- latest
:: dpcpp-ct -- latest
:: dpl -- latest
:: inspector -- latest
:: intelpython -- latest
:: ipp -- latest
:: ippcp -- latest
:: itac -- latest
:: mkl -- latest
:: modelzoo -- latest
:: modin -- latest
:: mpi -- latest
:: neural-compressor -- latest
:: pytorch -- latest
:: tbb -- latest
:: tensorflow -- latest
:: vpl -- latest
:: vtune -- latest
:: oneAPI environment initialized ::

steph@sjb-pc:$ conda update --all
Retrieving notices: ...working... done
Collecting package metadata (current_repodata.json): done
Solving environment: done

## Package Plan ##

environment location: /opt/intel/oneapi/intelpython/latest

The following packages will be downloaded:

package | build
---------------------------|-----------------
libabseil-20230125.3 | cxx17_h59595ed_1 1.2 MB file:///opt/intel/oneapi/conda_channel
libblas-3.9.0 |17_linux64_openblas 13 KB file:///opt/intel/oneapi/conda_channel
libcblas-3.9.0 |17_linux64_openblas 13 KB file:///opt/intel/oneapi/conda_channel
libgfortran-ng-13.1.0 | h69a702a_0 22 KB file:///opt/intel/oneapi/conda_channel
libgfortran5-13.1.0 | h15d22d2_0 1.8 MB file:///opt/intel/oneapi/conda_channel
liblapack-3.9.0 |17_linux64_openblas 13 KB file:///opt/intel/oneapi/conda_channel
libopenblas-0.3.23 |pthreads_h80387f5_0 9.4 MB file:///opt/intel/oneapi/conda_channel
libprotobuf-4.23.3 | hd1fb520_0 2.9 MB file:///opt/intel/oneapi/conda_channel
libstdcxx-ng-13.1.0 | hfd8a6a1_0 4.7 MB file:///opt/intel/oneapi/conda_channel
scipy-1.10.1 | py39h6183b62_3 18.4 MB file:///opt/intel/oneapi/conda_channel
------------------------------------------------------------
Total: 38.5 MB

The following NEW packages will be INSTALLED:

libblas conda_channel/linux-64::libblas-3.9.0-17_linux64_openblas
libcblas conda_channel/linux-64::libcblas-3.9.0-17_linux64_openblas
liblapack conda_channel/linux-64::liblapack-3.9.0-17_linux64_openblas
libopenblas conda_channel/linux-64::libopenblas-0.3.23-pthreads_h80387f5_0

The following packages will be UPDATED:

libabseil 20230125.2-cxx17_h59595ed_2 --> 20230125.3-cxx17_h59595ed_1
libgfortran-ng 12.2.0-h69a702a_19 --> 13.1.0-h69a702a_0
libgfortran5 12.2.0-h337968e_19 --> 13.1.0-h15d22d2_0
libprotobuf 4.23.2-hd1fb520_5 --> 4.23.3-hd1fb520_0
libstdcxx-ng 12.2.0-h46fd767_19 --> 13.1.0-hfd8a6a1_0
scipy 1.10.1-py39he5fd122_0 --> 1.10.1-py39h6183b62_3

Proceed ([y]/n)?

Downloading and Extracting Packages

Preparing transaction: done
Verifying transaction: failed

EnvironmentNotWritableError: The current user does not have write permissions to the target environment.
environment location: /opt/intel/oneapi/intelpython/latest
uid: 1000
gid: 1000

[Vasudha_Intel]

Yes, SUDO is required. 

[Vasudha_Intel]

Hello @bredsj, could you confirm if you have any further queries or concerns that require clarification? If not, we will proceed to close this issue.

[bredsj]

Good day

I have tried sudo conda update --all see result below:

steph@sjb-pc:$ sudo conda update --all
[sudo] password for steph:
sudo: conda: command not found
steph@sjb-pc:$

[Vasudha_Intel]

After installing the oneAPI AI Toolkits, If you need to update or install packages to the Intel environment consider cloning them to a separate environment first. This separation helps manage dependencies and avoids conflicts with existing installations. You can then make changes and test them in the cloned environment.

Also, the tools in the oneAPI toolkits are configured via environment variables. Set up your environment by sourcing the setvars script. You can find the setvars.sh script in the root folder of your oneAPI installation, which is typically /opt/intel/oneapi/

[Vasudha_Intel]

Hi, could you please confirm whether the approach of making changes and testing them in the cloned environment worked for you?

[bredsj]

No, I am not really sure what to do with cloning. I have already shown in one of my previous posts how the variables are set/sourced automatically through .bash.rc, again shown below for convenience.

[Vasudha_Intel]

Hi, thanks for replying, according to your issue you were successfully able to update Python in the Conda base environment however it was not updated within the intel environment. Please correct, if my understanding is incorrect.

[Vasudha_Intel]

Hi bresj,

There might not be write permissions for the pre-installed environments hence you could face this issue "EnvironmentNotWritableError".

These are a few commands to create new environments in the user directory so that you can install any other packages without any issues. This is a workaround for intel-aikit-pytorch environment.

Step 1: Clone the pre-install environment in a separate environment.

conda create -n aikit-pt -c intel intel-aikit-pytorch

Step 2: Activate the respective environment.

conda activate aikit-pt

Step 3: Update

conda update --all

Similarly, you can work with other pre-installed, environments.

[Vasudha_Intel]

Hi @bredsj,

Could you please confirm if the steps provided resolved your issue or if you're still experiencing any difficulties?

[Vasudha_Intel]

Hi @bredsj, we are closing this issue, if you face any other issues, please feel free to open another ticket, Thanks.

[bredsj]

 Thank you, noted and thanks for your assistance, the cloning didn't work. If I do find a solution I'll post it.