Jupyter Interface for NanoLabo / Matlantis Collaboration

JupyterLab / Jupyter Notebook are browser-based environments for Python. Through the Jupyter Interface for NanoLabo, it is possible to exchange atomic structures between the Atomic Simulation Environment(ASE) and NanoLabo. Additionally, this interface allows for the handling of atomic structures directly within JupyterLab.

Furthermore, a versatile neural network force field, Matlantis™, developed by Preferred Computational Chemistry (a joint venture between Preferred Networks and ENEOS), is available as a cloud service through Jupyter Notebook. By leveraging the Jupyter Interface for NanoLabo, users can model atomic structures in NanoLabo and execute simulations in Matlantis™. Additionally, it is possible to visualize the simulation results from Matlantis™in NanoLabo. This setup facilitates seamless collaboration between the two platforms.

Note

A separate optional license distinct from NanoLabo itself is required to use this feature.

Setting

Click on the icon mainmenuicon located in the upper left corner of the screen, and open Properties ‣ Jupyter Lab.

  • Type of Jupyter Lab

    Choose the default environment for the Jupyter Interface. Select Matlantis to use Matlantis™. For other JupyterLab environment, such as on-premises ones, set it to On Premise.

  • Launching Behavior

    Determine whether a dialog appears to select the environment when launching the Jupyter Interface. If you frequently switch between Matlantis™and on-premises environments and plan to use both, set this option to Ask Type of Jupyter Lab.

To access Matlantis™, initial log in is required. Click the button adjacent to Log-in label on the Matlantis tab for authentication. Once logged in, the Jupyter Notebook URL will be automatically populated in the URL field.

For on-premises environments, configure the connection to JupyterLab by setting the URL (e.g., http://localhost:8888/lab) and the Token in the On Premise tab.

The settings below apply to both tabs.

  • Uploading Path

    Configure the path for uploading atomic structure from NanoLabo to JupyterLab. Specify a relative path to the root directory of Jupyterlab.

  • Uploading File Type

    Specify .POSCAR format to include information on whether atomic coordinates are fixed/mobile during the uproad.

  • Using NGLView

    Include codes snippets for visualizing atomic structure with NGLView when importing files.

  • Drag & Drop

    Determine whether files dropped onto the screen are directly transferred to JupyterLab or processed by NanoLabo.

When the setting is completed, click the Python icon pythonicon located in the upper right corner of the window to verify if the JupyterLab screen is displayed.

NanoLabo → JupyterLab

To import an atomic structure from a project, click on Jupyter Lab from the projectmenuicon located in the lower left corner of the screen.

The atomic sturecture file is uploaded to JupyterLab, and simultaneously, the JupyterLab tab opens, displaying the Python code needed to import the atomic structure. Copy this code and execute it in JupyterLab to create the Atoms object using the ASE.

../_images/jupyter_import.svg

Hint

You can install NGLView by pip command.

$ pip install nglview

If you see an error like

AttributeError: 'super' object has no attribute '_ipython_display_'

during the visualization by NGLView, it may be resolved by specifying the version of ipywidgets .

$ pip install 'ipywidgets>=7.6.0,<8'

You can import not only from the project menu but also through the right-click menu of the atomic structure viewer or the project icon.

../_images/jupyter_context.svg

Hint

For instance, after creating a molecure using the Drawing Molecule feature, you can promptly upload it to JupyterLab from the right-click menu of the atomic structure viewer.

JupyterLab → NanoLabo

Initially, export the atomic strucure to a file in JupyterLab. Select a file format compatible with NanoLabo. Here is an example of exporting to the .POSCAR format.

from ase.io import write
write('export.POSCAR', myAtoms)

When you select Download from the right-click menu of a file in the file browser, it will be downloaded to NanoLabo, and a preview dialog will appear. Click Open to open the file as a project.

../_images/jupyter_export.svg