luseepy

About

A set of python utilities for performing various LuSEE-Night related calculations.

Documentation

This software is documented on the "Read the Docs" pages

Cached Data

There are datasets stored in the LuSEE-Night Google Drive.

Docker

For details, see the README file in the docker folder. The current setup involves building the luseepy image based on the refspec image. The latter contains C++ based software interfaces by means of the cppyy package.

Developing

To develop on your laptop, the easiest thing is to use the latest docker environment. Please install docker and pull the image

docker pull lusee/lusee-night-unity-luseepy:1.0 # or other appropriate version.

Next, checkout the lusee repo

git clone [email protected]:lusee-night/luseepy.git

Next, source the setup_env.sh script inside luseepy dir and even better, put it into your .bashrc.

source setup_env.sh

Now you have 4 utility functions:

• lpython starts runs python using shipped luseepy (unless one is in current dir)
• lpython_dev starts runs python using git checkout luseepy
• ljupyter starts jupyter notebook using shipped luseepy on port 9500
• ljupyter_dev starts jupyter notebook using git checkout luseepy on port 9600

To better understand the settings, it's worth it to take a look at the contents of setup_env.sh. In the current version, the HOME directory is mounted in the container, so it's possible to develop against the full current version of the luseepy/refspec suite. For example, you can now try running

# Simple calendar test
lpython tests/LunarCalendarTest.py
# to start jupyter
ljupyter

and then connect to the address given in the terminal output.

Environment variables

User is expected to set up the following environment variables:

• LUSEEPY_PATH -- path to the luseepy checkout
• LUSEEOPSIM_PATH -- path to the lusee opsim package (if used).
• LUSEE_DRIVE_DIR -- path to the checkout of the LuSEE-Night Google Drive

The following environment variables are set up by the setup_env.sh script:

• LUSEE_IMAGE -- docker image that has everything to run lusee

Singularity

NB. The example below corresponds to an early verion of software, and reference to the image below is deprected.

The tests folder contains CI-related and other testing scripts. Here's an example of a simple test run with Singularity, on a SDCC/BNL node, from the luseepy folder:

singularity exec -B /direct/phenix+u/mxmp/projects/luseepy --env PYTHONPATH=/direct/phenix+u/mxmp/projects/luseepy docker://lusee/lusee-night-foundation:0.1 ./tests/LunarCalendarTest.py

Cutting a new version

Cutting a new version entails:

• having a clean (non dev) version in __init__.py
• updating setup_env.sh
• tagging the github
• making new docker image
• bumping version again in __init__.py to a +0.1 and a dev

Any small fixes after the fact should be cumping version by 0.01. Large changes that break API should bump version into next integer.

Starting with simulations

Go to simulation sub-directory. Make sure the $LUSEE_DRIVE_DIR points to the stuff from the LUSEE Drive that the simulations needs (ULSA maps, beam). Run a short simulation as

python driver/run_sim.py config/realistic_example.yaml

In the same directory, open a jupyter notebook and plot the results for the NE cross-correlation, imaginary part as:

import lusee
D=lusee.Data('output/sim_output.fits')
plt.imshow(D[:,'01I',:],aspect='auto',extent=(D.freq[0], D.freq[-1],len(D.times),0))
plt.colorbar()
plt.xlabel('frequency (MHz)')
plt.ylabel('time number')