BoxShape

This class represents a geometric shape in the form of a box. It can be used to define a collection boundary for forwards transported gamma-rays, or as a generation surface in the backward case.

Constructor

class BoxShape(size, center=None, rotation=None)

This function creates a box with specified dimensions, in cm. The size can be given as a float (for a cube) or as a size-3 sequence of floats, indicating the width (along x, y) and height (along z) of the box. By default, the box is axis-aligned and centered on the origin. The centre argument can be used to specify an offset, in cm. Additionally, a 3x3 rotation matrix can be specified.

For example, the following creates a cubic box of 1m width with a 5 cm offset along the $z$-axis

>>> shape = goupil.BoxShape(
...     1e2, # cm
...     center = (0.0, 0.0, 5.0) # cm
... )

Attributes

BoxShape.center: numpy.ndarray

The Cartesian coordinates of the box center, in cm, as a size-3 numpy.ndarray.

BoxShape.rotation: numpy.ndarray | None

An optional 3x3 rotation matrix specifying the box orientation, if not axis-aligned.

BoxShape.size: numpy.ndarray

The box dimensions along the x, y and z-axis (before any rotation), in cm.

Methods

BoxShape.distance(states, /, *, reverse=None) → numpy.ndarray

Returns a numpy.ndarray of floats that indicate the distance of Monte Carlo states to the box surface, along their respective momentum directions.

By default, Monte Carlo states are assumed to propagate forwards. To trace states backwards instead, set the reverse flag to False.

BoxShape.inside(states, /) → numpy.ndarray

Returns a numpy.ndarray of booleans that indicate whether the provided Monte Carlo states are inside the box or not.

BoxShape.sample(states, /, *, engine=None, side=None, direction=None, weight=None)

Samples Monte Carlo states over the box surface. The weight boolean flag indicates wether the Monte Carlo states should be weighted by the inverse of the sampling PDF or not.

The side arguments refers to the surface side on which positions are sampled, within a numerical epsilon. Possible values are "Inside" or "Outside".

The direction arguments indicates the orientation of the sampled Monte Carlo states with respect to the box surface. Possible values are "Ingoing" or "Outgoing". By default, Monte Carlo states are considered as ingoing.

If a TransportEngine is provided (using the engine argument), then the sampling is configured based one the simulation mode. In a a forward simulation Monte Carlo states are generated on the inside by default, while in the backward case, the outside is used.

Note

At return from the sample method, the positions and directions of Monte Carlo states are modified in-place, along with their weights if applicable.