Skip to content

simonsobs/psplay

Repository files navigation

A tool to visualize and play with CMB maps. psplay is a jupyter extension to allow interactive visualization of CMB maps through Leaflet library. psplay also provides a set of tools based on pspy to compute and to show CMB power spectra. You can have a better idea of what psplay can do by watching this short video.

All the specific javascript library developed for psplay is done by Sigurd Naess. The build process and javascript architecture is highly inspired by ipyleaflet project.

image

image

Examples

Installation

To install, you will need to have or to install numpy. Then, run

$ pip install psplay [--user]

psplay runs within a Jupyter notebook and we recommend to use JupyterLab to fully enjoy psplay

To enable the extension within Jupyter, you will need to install several JupyterLab extensions especially the Python plotly library and the Jupyter widgets. You will also need nodejs library and its package manager npm. For JupyterLab > 2.0.0, you can do

$ jupyter labextension install plotlywidget jupyterlab-plotly
$ jupyter labextension install @jupyter-widgets/jupyterlab-manager @jupyter-widgets/jupyterlab-sidecar

Finally, you need to install the jupyter-leaflet-car extension

$ jupyter labextension install jupyter-leaflet jupyter-leaflet-car

Using docker

Given the number of requirements, you can use a docker image already made with the needed libraries and everything compiled and installed. You should first install docker for your operating system.

Then, you can run the simonsobs/psplay image by doing

$ docker run --rm -p 8888:8888 simonsobs/psplay:latest

After pulling the docker image, a JupyterLab instance should start. If your web browser does not automatically load the web page, you can copy-paste the JupyterLab URL.

You can bind a local directory to make it available within the docker container with the -v flag (see https://docs.docker.com/storage/bind-mounts for more details).

Installation from sources

For a development installation (requires npm)

$ git clone https://github.com/xgarrido/psplay.git
$ cd psplay
$ pip install -e .

If you are using the classic Jupyter Notebook you need to install the nbextension:

$ jupyter nbextension install --py --symlink --sys-prefix psplay
$ jupyter nbextension enable --py --sys-prefix psplay

If you are using JupyterLab, you need to install the labextension:

$ jupyter labextension install @jupyter-widgets/jupyterlab-manager js

Note for developers:

  • the -e pip option allows one to modify the Python code in-place. Restart the kernel in order to see the changes.
  • the --symlink argument on Linux or OS X allows one to modify the JavaScript code in-place. This feature is not available with Windows.

For automatically building the JavaScript code every time there is a change, run the following command from the psplay/js/ directory:

$ npm run watch

If you are on JupyterLab you also need to run the following in a separate terminal:

$ jupyter lab --watch

Every time a JavaScript build has terminated you need to refresh the Notebook page in order to load the JavaScript code again.

Authors

  • Xavier Garrido
  • Thibaut Louis
  • Sigurd Naess

The code is part of PSpipe the Simons Observatory power spectrum pipeline.