RESPRESSO homepage


What is it?

RESPRESSO (Rapid and Efficient SPectrum calculation based on RESponSe functiOn) is a novel power spectrum calculation method based on the response function. It consists of 2 components:

Starting from the simulation data, it utilizes the response function as a natural interpolator in the cosmological parameter (or model, more generally) space. Fast and accurate, it is ideal for MCMC analyses to constrain cosmological parameters using cosmic fluctuations on quasi-nonlinear scales. You can find more detail in our paper.


Python package documentation

The RESPRESSO python package is available from the link below. It works for python 2.7. After unzipping the tar.gz file, you can find the module "respresso" in it. You can place it in a directory with an appropriate path. You can also find some demonstrations of the code in the jupyter notebook format. It includes the following examples:

In addition to the package "respresso", the tar.gz file also contains an example python code "recon.py" and some test data. The example code works on the command line if you have python 2.7 installed properly: it simply accepts parameters and file names from the standard output and writes a result in a specified file. It exactly does the same 4 types of reconstructions as those in the above notebook. See README for its usage, which is also attached.

Advanced usages, such as user-defined number of steps (already implemented in the code) or reconstruction starting from user's own simulation templates (will be implemented) will be explained in a future release.


Versions


Conditions

RESPRESSO is a free software: you can redistribute it and/or modify it at your own risk. This program is distributed in the hope that it will be useful in scientific research, but without any warranty. If you use RESPRESSO for scientific work, we kindly ask you to cite the following paper:

Takahiro Nishimichi, Francis Bernardeau and Atsushi Taruya, Phys. Rev. D 96, 123515 (2017) (arXiv:1708.08946)


Acknowledgment

Numerical simulations from which the template nonlinear spectra are measured were performed on Cray XC30 cluster in Center for Computational Astrophysics (CfCA), National Astronomical Observatory of Japan (NAOJ). We thank the anonymous refree of the paper for pointing out the issue of the code regarding the unpike function that might appear in some versions of python 2.7.

last update: 14 Jan 2018 (Sun)

© Takahiro Nishimichi