Coordinate transformation

class arepy.coord.transf(show=False, **opt)

Coordinate transformations

Parameters

• show (bool) – Print out transformation information

• region (arepy.coord.regionBox or arepy.coord.regionSphere or arepy.coord.regionCone) – Coordinate region

• origin ([float]*3) – New coordinate origin

• align ([float]*3) – Vector of a new z-axis

• flip ([int]) – New order of the axis

• rotate ([float]) – Euler rotation angles

There is a standard sequence of transformations that can be used during the initialization of this class:

1. select - Select an initial spherical region from coordinates

2. translate - Set a new origin of the selected coordinates (e.g. center of the sphere)

3. align - Align the z-axis with some given vector (e.g. angular momentum)

4. flip - Flip axes (e.g. exchange y and z axes)

5. rotate - Add some additional rotation to the region

6. crop - Select a final region shape (e.g. box)

A simple initialization will look like:

import arepy as apy
import numpy as np

transf = apy.coord.transf(
    region = apy.coord.regionBox([0.2,0.8,0.2,0.8,0.2,0.8]),
    origin = [0.5,0.5,0.5],
    align = [0.3,0.5,0.0],
    flip = [0,2,1],
    rotate = [0,0,np.pi/5],
)

However, it is always possible to set an arbitrary list of transformations:

transf = apy.coord.transf()
transf.addSelection('sphere', apy.coord.regionSphere([0.2,0.8,0.2],0.8) )
transf.addTranslation('shift', [0.2,0.8,0.2])

convert(name, coord, **opt)

Perform selected transformations

Parameters

• name (str or list[str]) – Name or names of the transformations

• coord (list[[float]*len(dims)]) – List of coordinates/vectors

• dims (list[int]) – Dimensions of the vectors (default is [0,1,2])

Returns

Transformed coordinates

Initialized transformations can be applied on the set of coordinates:

coord = np.random.rand(20,3)
coord = transf.convert(['select','flip','crop'], coord)

It is also possible to convert vectors with an arbitrary set of dimensions. In the following case we are transforming the limits of a box region, where the vector dimensions are [x,x,y,y,z,z]:

limits = [0.2,0.8,0.2,0.8,0.2,0.8]
limits = transf.convert(['select','flip','crop'], limits, dim=[0,0,1,1,2,2])

select(name, data)

Select data corresponding to the coordinate region

Parameters

• name (str) – Name of the region transformation

• data – Data with the same length as the initial array with the coordinates

Returns

Data selected for the particles within the region

This function is used in the case when we have an additional dataset that corresponds to the coordinates (e.g. masses of the particles) and we want to additionally select its corresponding values:

coord = np.random.rand(20,3)
masses = np.random.rand(20)
coord = transf.convert(['select','flip','crop'], coord)
masses = transf.select('crop',masses)

addSelection(name, **opt)

Add selection

Parameters

• name (str) – Name of the transformation

• region (arepy.coord.regionBox or arepy.coord.regionSphere or arepy.coord.regionCone) – Coordinate region

addTranslation(name, origin)

Add a translation

Parameters

• name (str) – Name of the transformation

• origin ([float]*3) – Coordinates of the new origin

addAlignment(name, vector)

Add an alignment

Parameters

• name (str) – Name of the transformation

• vector ([float]*3) – Vector of a new z-axis

addRotation(name, angles)

Add a rotation

Parameters

• name (str) – Name of the transformation

• angles ([float]) – Euler rotation angles

addFlip(name, axes)

Flip axes

Parameters

• name (str) – Name of the transformation

• axes ([int]*3) – New order of the axes

show()

Print out all transformation into the command line