Follow the relevant walkthrough to create a new JupyterLab server.

This is a blank server. Follow the wizard to create your first personal server.

Make sure to select the environment module you wish to use, in this case the previously created "Python-Data-Science" module.

Then "Start" the server. This will allow you to access the "binaries" folder, which is necessary for installing dependencies.

Note that the server is running by the red "Stop" button on the server card.

After it has been started, navigate to the righthand sidebar and launch it. You can launch it by clicking the "Launch" button.

Once JupyterLab opens, click the module tab to make sure the given module is available.

From here, open a terminal to install dependencies in the correct binaries folder via conda commands.

Navigate to the modulefile folder at `/opt/modules/modulefile/<username>/Python-Data-Science/0.1.0.lua`

The modulefile should look something like this, and the binaries folder location will be shown as the "module_path"

Next, navigate to the `/home/jovyan/work` folder and create an empty environment.yaml file with `touch environment/yaml`

Then edit this file with `vi` or `nano`, providing the name, channels, and dependencies according to the conda environment you desire.

Finally, install the dependencies by running the conda command and using the path to the corresponding binaries folder, \`conda env create --prefix /opt/modules/binaries/6765eccfbc0456002a8f0dd0 --file environment.yaml\`

Once completed, you will be able to either activate the environment manually via the conda activate command or via the lefthand module tab by clicking the `load` button as noted previously.

Alert! Once the environment builds, rename the kernel display name so the new kernel won’t clash with existing one. Otherwise the kernel may be unavailable for use. The following information includes instructions on how to do this.

To view the kernel display name, navigate to the location of your kernel, in this case `/opt/modules/binaries/6773021af62105002a228ecb/share/jupyter/kernels`

In this case, the kernel folder name is `python3`. It is optional to rename this. Within this folder, you will find the `kernel.json` file which contains kernel metadata.

Run a jq command to replace the display name:

Key points:

- replace the display name with your environment name (Python-Data-Science)
- replace the binary ID with your own (`6773021af62105002a228ecb`)
- replace the kernel folder name with your own (python3)

`jq '.display_name= "Python-Data-Science"' /opt/modules/binaries/6773021af62105002a228ecb/share/jupyter/kernels/python3/kernel.json > tmp && mv tmp /opt/modules/binaries/6773021af62105002a228ecb/share/jupyter/kernels/python3/kernel.json`

The file will now look something like this.

You can now perform actions such as sharing this environment module with other users or using it as a dependency for a dashboard via Notebooks Hub.

Install Conda Dependencies on your Environment Module.