PyCharm Professional


I'm a very happy user of the IDE PyCharm Professional. If you do a lot of Python development, it's an ideal environment. It has a very wide range of the tools that are needed to develop, test, deploy, and manage. There are many useful plugins as well. Docker integration is also available.

The use of PC catches undefined symbols immediately rather than at run time. It also can enforce PEP8 style guides and perform quite rigorous code inspections.

Debugger support with vmprof is excellent.

As another example, my work requires running the same python code in multiple environments same as different machines. PC has tools for deployment via a number of secure connections. I use SFTP with SSHAgent most of the time.

The developers are very quick to respond to bug reports or queries.

Worth a look!

Talk on Low Calibration at RonFest

Talk on the infeasibility of SKA1-Low calibration at RonFest

Melanie kindly found a slot for me to talk about the calibration of SKA1-LOW for EOR imaging. Here is the presentation.

Followup to Memo 26


SDP has released my followup memo to memo 26. This one is SDP memo 27. I've addressed comments and criticisms, corrected some errors, and invented a new, more efficient approach to ionospheric calibration. The conclusion remains the same: SKA1-LOW cannot be calibrated to the level required for EOR imaging within the lifetime of the telescope.

Going further than this would be a good PhD topic! I don't intend to do much more.

RonFest


I'm currently attending RonFest - a celebration of Ron Eker's career and his 75th birthday. Ron was a very important mentor to me and countless other scientists. Here is my presentation on the early days of the VLA where Ron was director: "Learning how to do large-N radio interferometry".

FITS viewing in osx


I just ran across an integration of a fits viewer into OS X quick look. The quick look is a display of the FITS cards, and on demand it opens up QFitsView which is reasonably quick and intuitive image display app.

https://github.com/onekiloparsec/QLFits

Quite nice.