Opening Access to Fast Radio Interferometric Imaging¶
a.k.a. A Quick Demo of rtpipe¶
My science interests lie in fast radio transients, such as "fast radio bursts". But, frankly, a big part of why I enjoy studying these things is that it motivates a lot of fun and challenging work with software, algorithms, and data.
One example of that is our ongoing effort at the Very Large Array to do "fast imaging" --- forming images at millisecond cadence --- for massive transient surveys. Massive in this context is both in time (we've conducted a 200-hour survey and are in the midst of another) and data (1 hours on sky => 1 TB of data!). Doing this right requires a lot of custom software development, including the library I'm demoing here:
rtpipe is a library for radio interferometric data analysis that combines single-dish concepts like dedispersion and filters with interferometric concepts like images, the uv-plane, etc.. But it is probably more generally useful if you are interested in custom radio interferometric data analysis. If you are, then I highly recommend browsing the github contributions and blog of Peter Williams. Peter contributed both code and advice that helped shape
rtpipe (the mistakes are mine, though).
So, on with the demo...
Let's say you download one of the many public, TB-scale, VLA fast imaging data sets from the NRAO archive.
rtpipe lets you read and visualize that data in python, but of course most importantly, it lets one search it for fast radio transients.
import rtpipe.RT as rt import rtpipe.parsesdm as ps import rtpipe.parsecal as pc import rtlib_cython as rtlib from bokeh.plotting import figure, show from bokeh.io import output_notebook, gridplot output_notebook()