SelfCalibration

A project on survey design for precise self-calibration.

Authors:

• Rory Holmes (MPIA)
• David W. Hogg (NYU)

License:

Copyright 2012 the authors. All rights reserved.

If you want to license this code for use or re-use, get in touch.

Contributors:

• Hans-Walter Rix (MPIA)

Dependencies:

• numpy
• matplotlib
• scipy

Running the code:

A self-calibration simulation can be run by calling the simulation.run_sim(parameter_dictionary) function, where the parameter_dictionary contains all the parameters for this simulation run. This functions controls a full self-calibration simulation: it generates the synthetic sky, surveys the sky according to the specified survey strategy and then self-calibrates the resultant dataset. This function then returns the performance of the self-calibration procedure. See default_parameters.py for all of the required simulation parameters. Running the plot.py script on the output directory will produce plots from this self-calibration run. The package has been built in this way to allow for simple multiprocessing, as the simulation.run_sim(parameter_dictionary) function can be called multiple times in parallel with different parameter dictionaries. For example, to test how the self-calibration procedure is sensitive to the number of sources, one can simply run this function with multiple parameter dictionaries in which the number of sources parameter is varied. A number of sample scripts have been included to show how this package can be used.

Modules:

This package contains seven modules for the self-calibration simulations:

• analysis.py contains the functions used to analyze the fitted instrument response relative to the true instrument response.
• save_out.py contains the functions that save out all the data from the simulation. NB that data_dir parameter in the parameter dictionary must be specified, otherwise no data is saved out from a simulation run.
• self_calibration.py contains the self-calibration procedures.
• simulation.py contains the master function that controls a full the full self-calibration simulation.
• survey.py contains the functions used to perform a survey of the synthetic sky, such as the imager measurement model.
• transformations.py contains useful transformation functions that are used throughout the code.
• true_functions.py contains all the functions representing the true parts of the simulations, such as the actual instrument response that is fitted for in the self-calibration simulations.

Plotting:

Plots for a simulation run can be generated by: plot.py [output_directory].

Parameters:

A complete parameter dictionary must be past to simulation.run_sim() for each simulation run. The code does not check for missing parameters before beginning the simulations. See default_parameters.py for a full list of required parameters.

Known issues:

• ...

Migration from svn:

Hogg migrated this from svn with something like

cd
git svn clone svn+ssh://astrometry.net/svn/trunk/projects/euclid/ --no-minimize-url --authors-file ~/authors
cd euclid
git remote add origin git@github.com:davidwhogg/SelfCalibration.git
git push origin master