sophipkorte / so-lenspipe

SO L3.1 pipeline scripts and libraries for CMB lensing reconstruction and auto-spectra

solenspipe

Pipeline libraries and scripts for L3.1

• Free software: BSD license

Dependencies

NERSC tip: you may need the python/3.7-anaconda-2019.07 module. You can add module load python/3.7-anaconda-2019.07 to your ~/.bash_profile.ext. Also, when running quick tests on the login node (e.g. to test imports after setting up), you should run export DISABLE_MPI=true since MPI calls do not work on the cori login node.

Here are all the pacakges you'll need before you can run this library and scripts therein:

• pixell (if running on NERSC, run python setup.py build_ext -i --fcompiler=intelem --compiler=intelem followed by adding the directory to your PYTHONPATH; else run python setup.py install --user); test by running py.test -s
• falafel (pip install -e . --user)
• tempura (Clone the repository then do python setup.py build_ext -i and then add to your PYTHONPATH)
• camb (pip install camb --user)
• orphics (pip install -e . --user)
• enlib (just need enlib/bench.py for benchmarking ; git clone the repo and add to PYTHONPATH)
• soapack (pip install -e . --user)
• pyfisher (pip install -e . --user)
• actsims (clone repo, checkout new_scheme branch, and then pip install -e . --user)
• Other miscellaneous packages: healpy, Cython, astropy, numpy, scipy, matplotlib, pyyaml, h5py, Pillow (Python Image Library), toml

Installing

To install, run:

python setup.py build_ext -i
pip install -e . --user

Then copy input/config_template.yml to input/config.yml and edit it to match paths on your system (specifically, the data_path variable in the .yml file will need to be changed to be for a directory of your own).

Demo

Run python examples/bias.py -h and if the installation is succesfull, you should see

usage: bias.py [-h] [--nsims-n0 NSIMS_N0] [--nsims-n1 NSIMS_N1] [--healpix]
                           [--new-scheme] [--lmax LMAX] [--lmin LMIN] [--biases BIASES]
                           version est1 est2

Verify and benchmark RDN0 on the full noiseless sky.

positional arguments:
  version              Version name.
  est1                 Estimator 1, one of TT,TE,EE,EB,TB,MV,MVPOL.
  est2                 Estimator 2, same as above.

optional arguments:
  -h, --help           show this help message and exit
  --nsims-n0 NSIMS_N0  Number of sims.
  --nsims-n1 NSIMS_N1  Number of sims.
  --healpix            Use healpix instead of CAR.
  --new-scheme         New simulation scheme.
  --lmax LMAX          Maximum multipole for lensing.
  --lmin LMIN          Minimum multipole for lensing.
  --biases BIASES      Maximum multipole for lensing.

For a test beyond the imports, you can run python examples/bias.py test TT TT --lmax 300 --nsims-n0 1 --nsims-n1 1 but you'll need some files in your data_path directory to get going.

Note that if working on NERSC, you might have to run the scripts on an interactive node.

Contributing

If you have write access to this repository, please:

1. create a new branch
2. push your changes to that branch
3. merge or rebase to get in sync with master
4. submit a pull request on github

If you do not have write access, create a fork of this repository and proceed as described above.