jupyter_rs_radia

Widget for visualizing 3D Radia models in a jupyter notebook

RadiaViewer allows users to render a magnet geometry and solve for its fields in a self-contained widget. The rendering is done by the VTK.js library.

NOTE: This depends on JupyterLab 3.x

Installation

To install use pip:

$ pip install .
$ jupyter labextension install .

For a development installation (requires npm),

$ git clone https://github.com/radiasoft/jupyter_rs_radia.git
$ cd jupyter_rs_radia
$ pip install -e .
$ jlpm install:extension

When actively developing your extension, use watch to hot reload on changes:

$ jlpm watch:labextension

This takes a minute or so to get started, but then allows you to hot-reload your javascript extension. To see a change, save your javascript, watch the terminal for an update.

Setup

This guide assumes familiarity with Raida and its python API.

1. In your notebook, add from jupyter_rs_radia import radia_viewer
2. Create an instance of the RadiaViewer: rv = radia_viewer.RadiaViewer()
3. Define your geometries, materials, etc.
4. Add the geometries you wish to display: rv.add_geometry(<name>, <ref>)
5. Display them! rv.display()

Understanding the viewer

You should see your selected geometry as below:

Using the mouse or trackpad

Click and drag in the 3D display to rotate the objects freehand.

Shift-click and drag to pan.

Scroll to zoom in and out.

Control-click to select objects and fields (see below)

Camera controls

Reset Camera: moves the camera back to the default position.

X, Y, Z: moves the camera so that axis is pointing in (⊙) or out (⦻) of the screen.

Show marker toggles the orientation cube in the upper right corner.

Viewer properties

Background color: pops up a color picker to change the background color of the 3D area.

Geometry properties

Geometry: dropdown menu of geometries added in step (4) above. Defaults to the 1st added

View:

• Objects: 3D visualization of the geometry (default)
• Fields: representation of various fields calculated by Radia (see below)

Object properties

Note: only visible when the View is set to "Objects".

Object color: first control-click an object in the 3D area to select it. Then click Object color to pop up a color picker. When complete, control-click again to deselect the object. If the selected geometry is a container, each member is individually selectable. However, if a member is also a container, it is not further subdivided

Surface alpha: fades the faces of the objects. The edges are unaffected.

Show edges: toggles the edges of the objects.

Field properties

Note: only visible when the View is set to "Fields".

Color map: colors field vectors according to their magnitude. Choices include:

• afmhot
• coolwarm
• grayscale
• jet
• viridis (default)

Scaling: changes the relative sizes of field vectors according to their magnitude. Choices include:

• Uniform: all vectors have the same length

• Linear: vectors scale linearly according to their relative magnitude, i.e.

   *s =  (|v| - minV) / (maxV - minV)*
  

  where minV and maxV are the minimum and maximum magnitudes among the vectors

• Log: vectors scale linearly according to the relative log of their magnitude, i.e.

    *s =  (maxV - minV) * (ln(|v|) - minLogV) / (maxLogV - minLogV)*
  

  where minLogV and maxLogV are the minimum and maximum log magnitudes. The (maxV - minV) factor is there to keep the linear- and log-scaled vectors in the same visual range

Radia controls

Field: selects the field plotted. Choices include:

• M (magnetization)
• B (magnetic field)
• A (vector potential)
• H (magnetic field strength)
• J (current density)

Radia calculates the magnetization within subdivided passive magnet elements. The other fields require the user to define where they are evaluated:

Path adds points for evaluation of the fields. Choices include:

• Line: set a start point, end point, and number of evaluation points.
• Circle: set the center, radius, euler angles of the normal of circle's plane, and number of evaluation points.
• Manual: add points one at a time
• File: upload points from a text file. The coordinates must be flattened and comma-delimited (i.e. x0, y0, z0, x1, y1, z1,...)

Add the path(s) of interest with the + button. The points will be appended to those already in place, with the exception of those added from a file. In that case any existing points are deleted. List all the current points with rv.get_field_points().

Precision, Max iterations, Method: refer to the Radia documentation for precise definitions of these settings.

Solve: execute RadSolve(). When complete, the solution will be reflected in the display of field vectors. Control-click an individual vector to see its magnitude and direction, or list the full set of field points and values wiith rv.get_result().