Skip to end of metadata
Go to start of metadata

You are viewing an old version of this page. View the current version.

Compare with Current View Page History

« Previous Version 8 Next »

On this page:



Conda is a tool commonly used to manage dependencies and projects. You can create a conda environment for each project and install the packages that you need in that conda environment. If you make changes to one of your conda environments, your other environments are not affected. For more information, please refer to the Conda User Guide.

The following examples illustrate common conda tasks using JupyterLab in IDAS. 

Example: Create a conda environment with a Python kernel 

1. First, log in to the IDAS research instance with Python. If you are a student in a class that uses Python in IDAS, follow the instructions here to access your class instance.

2. In JupyterLab, click the "Terminal" tile under "Other" to start a Terminal session.


3. In this example, we will create a conda environment with Python 3.8 and name the environment "py38".


conda create -n py38 python=3.8

When prompted "Proceed ([y]/n)?", press "y" to proceed.

Side notes:

    • You can also install packages when creating an environment, e.g.: 
      • conda create -n another-env python=3.8 numpy requests
    • You can specify the versions of the packages, e.g.: 
      • conda create -n another-env python=3.7 numpy=1.16.1 requests=2.19.1
    • You can also create a conda environment from an environment.yml file, e.g,
      • conda env create -f environment.yml
    • For more examples of creating conda environments, please see "Managing environments" in the Conda User Guide.

4. Once this conda environment has been created, we can activate it.

conda activate py38

5. The command prompt in your Terminal will change to indicate the active environment.

(py38) hawkid@idas-research-hawkid:~$

6. Next, we create a kernel in order to use Jupyter Notebook with this conda environment. Install the IPython kernel:

conda install ipykernel

When prompted "Proceed ([y]/n)?", press "y" to proceed.

7. Now install a kernel in this environment:

python -m ipykernel install --user --name py38 --display-name "Python Project Vis"

The value for "--name" is used by Jupyter internally. Any existing kernel with the same "--name" value will be overwritten. The "--display-name" will be displayed in the Notebook menu in the JupyterLab Launcher page.

8. Go back to the JupyterLab Launcher page by pressing Ctrl+Shift+L (Windows) or Cmd+Shift+L (Mac). Under "Notebook", a new option for your kernel will now be available. In this example screenshot, the "Python Project Vis" kernel was just installed and now became available to use.

Click on that new option to start a notebook. In that notebook, you can use the packages that you installed in the conda environment.




9. To install Python packages to this conda environment using a Jupyter notebook:

a. Start a new Jupyter notebook by selecting the Python kernel you just installed. In this example, that is the "Python Project Vis" kernel in the previous step.

b. In a cell in this new notebook, use the %conda magic that is built into IPython to install packages within the current kernel. For example, install the "pandas" package:

c. Then we can load and use the package as usual:


10. We can also install Python packages to this conda environment using Terminal:

conda install --name py38 pandas

Example: Create a conda environment with an R kernel 

1. First, log in to the IDAS research instance with Python. If you are a student in a class that uses Python in IDAS, follow the instructions here to access your class instance.

2. In JupyterLab, click the "Terminal" tile under "Other" to start a Terminal session.


3. In this example, we will create a conda environment with R and name the environment "r-env".

conda create -n r-env r-essentials r-base

When prompted "Proceed ([y]/n)?", press "y" to proceed.

Side notes:

4. Once this conda environment is created, we can activate it.

conda activate r-env

5. The command prompt in your Terminal will change to indicate the active environment.

(r-env) hawkid@idas-research-hawkid:~$

6. Next, we create a kernel in order to use Jupyter Notebook with this conda environment. We will install a kernel through IRkernel. The r-irkernel package was already installed with the r-essentials bundle earlier. You can check to make sure you have r-irkernel in your conda environment:

(r-env) hawkid@idas-research-hawkid:~$ conda list irkernel
# packages in environment at /home/hawkid/.conda/envs/r-env:
#
# Name                    Version                   Build  Channel
r-irkernel                0.8.15                    r36_0    defaults

7. Now install a kernel in R:

# launch R in Terminal
R
> library(IRkernel)
> IRkernel::installspec(name = "ir361", displayname = "R 3.6.1")

The value for "name" is used by Jupyter internally. Any existing kernel with the same "name" value will be overwritten. The "displayname" will be displayed in the Notebook menu in JupyterLab.

Note that we specified the R version (3.6.1) in the "name" and "displayname". This helps remind us what R version we are using, especially if we have multiple conda environments and kernels.


8. Go back to the JupyterLab Launcher page by pressing Ctrl+Shift+L (Windows) or Cmd+Shift+L (Mac). Under "Notebook", a new option for your kernel will now be available, with the name "R 3.6.1"

Click on that new option to start a notebook. In that notebook, you can use the packages that you installed in the conda environment.


9. To install R packages in this conda environment, in Terminal:

conda install --name r-env r-abind


10. Then you can load and use the package in Jupyter Notebook as usual. If you have an open notebook, you can use the package in the notebook right after you install the package in Terminal. You don't need to refresh the notebook.

11. In general, any packages that are installed in the conda environment will be available to use in the notebook. To view a list of installed packages in the conda environment, type in Terminal:

conda list -n r-env

12. To view a list of installed packages in Jupyter notebook, use the installed.packages() function:

Other useful conda commands 

The following commands may be useful to manage conda environments. Type the following commands in Terminal. For more information, please refer to the "Managing environments" and "Managing packages" sections in the Conda User Guide.

  • List all of your environments. In the output, your current environment will be marked with an asterisk (*).

    conda info --envs

    or

    conda env list

  • To activate a conda environment that was previously created, type (without the angle brackets):

    conda activate <environment-name>

  • Deactivate a conda environment once you are finished working with it:

    conda deactivate

    • Or, to return to the "base" environment, type (with no environment specified):

      conda activate
  • View a list of packages installed in an environment:

    • If the environment is not activated, type (without the angle brackets):

      conda list -n <environment-name>

    • If the environment was already activated with "conda activate" earlier:

      conda list

  • To see if a specific package is installed in an environment, type (without the angle brackets):

    conda list -n <environment-name> <package-name>

    • Or, if the environment was already activated with "conda activate" earlier, type (without the angle brackets):

      conda list <package-name>
  • To install a package in an environment:
conda install --name <environment-name> <package-name>

  • Remove an environment:

    conda remove --name <environment-name> --all

    or

    conda env remove --name <environment-name>

Using pip in a conda environment

Please note that issues may arise when using pip and conda together. The "Using pip in an environment" section in the Conda User Guide outlines best practices for using pip in a conda environment.


Contact

If you have any questions or comments, please contact research-computing@uiowa.edu.


  • No labels