Systematics For Cosmic Microwave Background (s4cmb), is a package to study instrumental systematic effects in the context of current and future Cosmic Microwave Background experiments.

The pipeline is mainly written in python and it has the following dependencies:

• numpy, matplotlib
• astropy, ephem, pyslalib, healpy (astro libs)
• f2py, weave (interfacing with python)

While we use python 2.7, we try to make it compatible with python 3.x. If you are using python 3.x and you encounter an error, please open an issue or a pull request so that we fix it asap.

Some parts of the pipeline are written in C (and compiled on-the-fly via the package weave), and in Fortran (to come). The latter is interfaced with python using f2py. The compilation is done usually when you install the package (see setup.py), but we also provide a Makefile for more customized compilations (see dir/Makefile).

You can easily install the package using pip

pip install s4cmb

Otherwise you can fork the repo from the github repository and clone it to your machine. Use the setup.py for the installation. Just run:

python setup.py install

Make sure you have correct permissions (otherwise just add --user at the end of the command). You can also directly use the code by updating manually your PYTHONPATH. Just add in your bashrc:

s4cmbPATH=/path/to/the/package
export PYTHONPATH=$PYTHONPATH:$s4cmbPATH

Then run the test suite and the coverage:

./coverage_and_test.sh

It should print the actual coverage of the test suite, and exit with no errors.

Alternatively if you do not want install the package on your computer, we provide a docker image for s4cmb with always the latest version. Install docker on your computer, and pull the image:

docker pull julienpeloton/s4cmb:latest

Then create a new container and run an interactive session by just running

docker run -i -t julienpeloton/s4cmb:latest bash

You can find notebooks describing how to use basic functionalities of s4cmb in a separate repository: s4cmb_notebooks . We also provide a quick end-to-end example for using the package with MPI. Try to run (you will need the package mpi4py)

mpirun -n <nproc> python examples/simple_app.py -inifile examples/simple_parameters.ini -tag test

where nproc should not be greater than the number of scans to run. Note that for NERSC users, we also provide a submission script for jobs on Cori (see examples/nersc_cori.batch).

Let's say we want to build an instrument, a scanning strategy, and scan the sky to obtain data. Say we also want to inject crosstalk between detectors, and then reconstruct the sky maps with the contamination.

• Step 1 [parameters initialisation]: create a ini file with your parameters. The best is to copy the one provided (examples/simple_parameters.ini) and change the values to yours. Do not forget to update the paths to data!

[s4cmb]
## Parameter file for a fake experiment.
## Run ID
tag = gros
name_instrument = fake

...

• Step 2 [start the App]: Create a python script, and import relevant modules

## python 2/3 compatibility.
from __future__ import division, absolute_import, print_function

## If you want to perform parallel computation.
from mpi4py import MPI

## Import modules and routines from s4cmb.
import s4cmb

...

• Step 3 [tell the App what to read]: link your inifile to your App. For that one we will use the module argparse for example. Also add any useful args you want to pass:

def addargs(parser):
    """ Parse command line arguments for s4cmb """

    ## Defaults args - load instrument, scan and sky parameters
    parser.add_argument(
        '-inifile', dest='inifile',
        required=True,
        help='Configuration file with parameter values.')

    ...

• Step 3 [load background]: Tell the App to load the background (instrument, scan, and so on).

if __name__ == "__main__":
    """
    Launch the pipeline!
    """
    <grab args>

    ## Initialise our input maps.
    sky_in = s4cmb.input_sky.HealpixFitsMap(...)

    ## Initialise our instrument.
    inst = s4cmb.instrument.Hardware(...)

    ## Initialize our scanning strategy and run the scans.
    scan = s4cmb.scanning_strategy.ScanningStrategy(...)
    scan.run()

• Step 4 [perform computations]: Loop over scans, and for each scan do map2tod -> inject crosstalk -> tod2map. Note that the maps are coadded on the fly so that sky_out_tot contains all scans.

for CESnumber in range(scan.nCES):
    tod = s4cmb.tod.TimeOrderedDataPairDiff(...)

    ## Initialise map containers for each processor
    if CESnumber == 0:
        sky_out_tot = s4cmb.tod.OutputSkyMap(...)

    ## Scan input map to get TODs
    d = np.array([
        tod.map2tod(det) for det in range(inst.focal_plane.nbolometer)])

    ## Inject crosstalk
    s4cmb.systematics.inject_crosstalk_inside_SQUID(d, ...)

    ## Project TOD back to maps
    tod.tod2map(d, sky_out_tot)

• Step 5 [write on disk your maps]: We provide some routines to write fits file but feel free to write your routines with your favourite I/O!

s4cmb.xpure.write_maps_a_la_xpure(...)
s4cmb.xpure.write_weights_a_la_xpure(...)

Et voilà! You can find this complete example in examples/so_crosstalk_app.py.

• Add WHWP demodulation module.
• Add correlated noise simulator (and update mapmaking weights).

• Julien Peloton (j.peloton at sussex.ac.uk)
• Giulio Fabbian (gfabbian at ias.u-psud.fr)