tmolteno / disko_array_opt

Array Optimization using the Singular Values of the telescope operator

Author: Tim Molteno

See the jupyter notebook for details.

INSTALL:

Issue the command

make install

RUN

Issue the command

make

or have a look at the runtime options by running the array_opt.py command as follows:

python3 array_opt.py --help

Sample output below

usage: array_opt.py [-h] [--output OUTPUT] [--iter ITER] [--nant NANT] [--narm NARM] [--arcmin ARCMIN] [--radius RADIUS] [--radius-min RADIUS_MIN]
                    [--spacing SPACING] [--fov FOV] [--learning-rate LEARNING_RATE]

DiSkO Array: Optimize an array layout using the singular values of the array operator

optional arguments:
-h, --help            show this help message and exit
--output OUTPUT       Root of output file names. (default: optimized_array)
--iter ITER           Number of iterations. (default: 100)
--nant NANT           Number of antennas per arm. (default: 8)
--narm NARM           Number of arms. (default: 3)
--arcmin ARCMIN       Resolution of the sky in arc minutes. (default: 120)
--radius RADIUS       Length of each arm in meters. (default: 2.0)
--radius-min RADIUS_MIN
                        Minimum antenna position along each arm in meters. (default: 0.1)
--spacing SPACING     Minimum antenna spacing. (default: 0.15)
--fov FOV             Field of view in degrees (default: 180.0)
--learning-rate LEARNING_RATE
                        Optimizer learning rate. (default: 0.02)

TODO:

• Add the ability to start from a JSON file.

The following array reached 16.1:

    Arm 0: [0.224 0.344 0.787 1.292 1.600 1.928 2.864 3.121]
    Arm 120: [0.229 0.897 1.324 2.250 2.585 2.712 3.544 4.174]
    Arm 240: [0.094 0.106 1.001 1.158 1.273 1.575 2.148 3.008]