The Python Podcast.__init__

Share on social media:

Visit our site to listen to past episodes, support the show, and sign up for our mailing list.

Summary

What is Solar Physics? How does it differ from AstroPhysics? What does this all have to do with Python? In this episode we answer all of those questions when we interview Stuart Mumford about his work on SunPy. So put on your sunglasses and learn about how to use Python to decipher the secrets of our closest star.

Brief Introduction

• Hello and welcome to Podcast.__init__, the podcast about Python and the people who make it great.
• Subscribe on iTunes, Stitcher, TuneIn or RSS
• Follow us on Twitter or Google+
• Give us feedback! Leave a review on iTunes, Tweet to us, send us an email or leave us a message on Google+
• I would like to thank everyone who has donated to the show. Your contributions help us make the show sustainable. For details on how to support the show you can visit our site at pythonpodcast.com
• I would also like to thank Hired, a job marketplace for developers, for sponsoring this episode of Podcast.__init__. Use the link hired.com/podcastinit to double your signing bonus.
• Linode is sponsoring us this week. Check them out at linode.com/podcastinit and get a $10 credit to try out their fast and reliable Linux virtual servers for your next project
• We are recording today on November 17th, 2015 and your hosts as usual are Tobias Macey and Chris Patti
• Today we are interviewing Stuart Mumford about SunPy

Interview with Stuart Mumford

• Introductions
• How did you get introduced to Python? – Chris
• Can you explain what the research and applications of solar physics are and how SunPy facilitates those activities? – Tobias
• What was your inspiration for the SunPy project and what are you using it for in your research? – Tobias
• Can you tell us what SunPy’s map and light curve classes are and how they might be used? – Chris
• Are there any considerations that you need to be aware of when writing software libraries for practitioners of the hard sciences that would be different if the target audience were software engineers? – Tobias
• Can SunPy consume data directly from telescopes and other observational apparatus? – Chris
• I noticed on the project site that SunPy leverages AstroPy internally. Can you describe the relationship between the two projects and why someone might want to use SunPy in place of or in addition to AstroPy? – Tobias
• Looking at the documentation I got the impression that there is a fair amount of visual representation of data for analysis. Can you describe some of the challenges that has posed? Is there integrated support for project Jupyter and are there other graphical environments that SunPy supports? – Tobias
• What are some of the most interesting applications that SunPy has been used for? – Chris

Picks

• Tobias
  • Elm
  • Avro
  • Common Sense Media



• Chris
  • Massdrop
  • 21st Amendment Fireside Chat
  • Extra Creditz



• Stuart
  • Live ISS Stream with space-to-ground radio
  • Live ISS HD video stream 24/7
  • yt
  • Calf Studio – Live Audio Processing

Keep In Touch

• Twitter(@sunpyproject)
• SunPy.org
• GitHub
• IRC

The intro and outro music is from Requiem for a Fish The Freak Fandango Orchestra / CC BY-SA

[00:00:14] Unknown:

Hello, and welcome to podcast. Init, the podcast about Python and the people who make it great. You can subscribe to our show on iTunes, Stitcher, TuneIn Radio, or add our RSS feed to your podcatcher.   You can also follow us on Twitter or Google plus and please give us feedback. Leave us a review on iTunes, send us a tweet, send us an email, leave a message on Google plus or leave a comment on our show notes.   Leaving a review on iTunes helps other people find the show so that they can enjoy it as much as you do.   I'd like to thank everyone who hasn't donated to the show. Your contributions help us make the show sustainable. For details on how to support the show, you can visit our site at pythonpodcast.com.  

I would also like to thank Hired, a job marketplace for developers, for sponsoring this episode of podcast.init.   Use the link hired.com/podcastinit   to double your signing bonus.   Linode is sponsoring us this week as well. You can check them out at linodes.com/podcastinit   and get a $10 credit to try out their fast and reliable Linux virtual servers for your next project.   We are recording today on November 17, 2015, and your hosts as usual are Tobias Macy and Chris Patti.   Today, we are interviewing Stuart Mumford about SunPi. Stuart, could you introduce yourself, please?  

[00:01:24] Unknown:

Hi.   I, just finished writing up my PhD on, numerical solar physics University at Sheffield in the UK.   And for the last couple of years, I've also been the lead developer of the Sunpipe project.  

[00:01:39] Unknown:

So, Stuart, how did you get into programming and Python in particular?  

[00:01:43] Unknown:

So I   my first, like, kind of formal introduction to programming was   a Fortran 90 course,   in my physics undergrad.   We did I had a few modules during my undergraduate on,   FORTRAN programming.   Sometime late in my 2nd year of my degree, my lab partner introduced me to Python.   We taught ourselves it by doing all our data reduction for our 3rd year lab experiments in   Python. And by the end of my 3rd year, I had a relatively good handle on scientific program.  

[00:02:19] Unknown:

And can you explain what the research and applications of solar physics are and how Sun Pi facilitates those activities?  

[00:02:27] Unknown:

So the part of solar physics that I   may mainly am involved with and Sun Pi kinda relates to is the   part solar physics that studies the interaction of the sun and the earth.   Specifically,   all of the dramatic events that   affect the earth in 1 way or another, be it   potential to knock out satellites or power grids, kind of the large space weather events are all   all kinda originate in the outer layers of the solar atmosphere.   And the main focus of   most solar physics research is the heating of the solar atmosphere, especially the outer layers of the solar atmosphere   heated to 1, 000, 000 of degrees.  

And   the   precise mechanism of how that energy is transported up in the solar atmosphere isn't known. And   so this   the application of most solar physics would be that   by studying this heating,   we would have a better handle on the physics of what causes these events,   all space weather, and there are all effects on it.   SunPy is a library that is   designed to,   represent and   allow scientists to process   observational data of the sun. So   be it   images, spectral data, or time series information.  

[00:03:52] Unknown:

Can that data also be applied for research in terms of   solar renewables, like solar electricity or solar heat?  

[00:04:03] Unknown:

Maybe a little bit, but not really. Most of   that sort of most of the solar side of that, to my knowledge, is about   how   much   radiance of the   sun lands on certain places and stuff, and that's not it's more kind of direct of sun pie is more direct observational   data.  

[00:04:25] Unknown:

And what is the relation to between solar physics and astrophysics?   Are there any   differences in the problem domains or the ways that the research is conducted?  

[00:04:38] Unknown:

There are an awful lot of similarities   between solar physics and,   like, night side astronomy or dark side astronomy.   So the physics got solar astronomy is really a subset.   Sort of a basic level, they do very similar things. Like, they're both observing   space,   through a telescope onto a CCD.   You   capture light and form an image   and then   have to worry about coordinate systems or transformations   of, like, kind of distortions through your telescope or the atmosphere. There's an awful lot of similarities.   There are a lot of differences as well, especially around   sort of the atmospheric effects are different because   you're not observing point sources in solar physics, and you end up with an awful lot more light because you're observing something that's much closer.  



[00:05:33] Unknown:

It's very interesting. It's   definitely a fascinating area of research and 1 that I would like to be able to spend more time on given actually having some free time available. So what was your inspiration for the Sun Pi project, and what are you using it for in your research?  

[00:05:48] Unknown:

So I only got involved with Sun Pi   just after its 1st naught 0.1 release.   I actually saw the release announcement   for SunPi naught 0.1 on Twitter on my way into work. And,   we were excited thinking, oh, someone started writing that thing that I've always said I was gonna do when I had a free year.   And I got into the office and logged into the IRC channel and managed to worm my way into leading the whole project.   SunPi,   the motivation for SunPai is   to provide a more modern toolset for solar physicists.   An awful lot of solar physics at the moment uses a proprietary   language called IDL,   which has   some rather major flaws.  

But it's also very   it's, like, it's obviously very useful for solar physics because there's a massive collection of code called solar soft, which is   really a collection of the last kind of 10, 15 years' worth of   primarily   space based   instrument pipelines. There's about 4, 000, 000 lines of IDL code in the SodaSoft package.   The   purpose of SunPy is to provide   the sort of core tools that exist in that library,   not the   instrument calibration routines, which were written by the people that built instruments.   Sunby   the goal of Sunby is to provide a modern Python based   way of representing   calibrated solar data   for people to work on and do scientific study with.  



[00:07:40] Unknown:

So can you tell us what SunPy's map and light curve classes are and how they might be used?  

[00:07:46] Unknown:

So SunPy   provides   3 core   sort of data types,   map   for   currently 2 d image data,   light curve for time series, which is based on the excellent Pandas library,   and a spectra class for holding   spectra data and,   data which is wavelength and time. The idea of these is they provide the base framework for all the other code to kind of work with.   There are   quite a few   different like, SunPi provides access to lots of different instruments,   lots of different data sources.   And the idea of these data types is that they kind of provide   a unified interface to   a wide variety of different   data sources and instrument types. So the 1 I've done the most work on is the map class, which is   currently limited to 2 dimensional data,   plus collections thereof.  

And   the   the functionality it provides is primarily   a way   of holding   the number array, which is the image and all of the metadata that comes   with that   image from wherever that source is. I mean, normally, it's in a format called FITS,   and   we read that FITS file and extract a NumPy array and a dictionary, basically, which is the data and the metadata.   Then the map class   provides   a standard interface to those 2 things that account for some instruments   have,   say,   the name of the instrument in a with a key in the dictionary of   instrument, and some of them have truncated it to instrument.  

So the you can access the instrument by doing,   map dot instrument, he says, off top of his head, hoping that's right.   But and it doesn't matter how that information   arrived to you   in the file. You   so if you written a code for 1 instrument   and you have accessed the name of the instrument and the size of the array in physical units,   If you can swap that out for   a different   instrument's data, and your code will still work fine.   The other main thing that map does is deal with   providing transformations   from   array coordinates, so array indexes,   to physical coordinates. So   there are 2 or 3   solar,   like, solar coordinates, which   so solar coordinate systems that represent   points on the sun,   be it points on an image on the sun or 3 d   based around the center of the sun coordinates.  

What map does is convert from   the array indexes to   the helioprojective   coordinate system, which is   measured in, like, angular distance from the center of the solar disk. And that primarily   uses Astropy   to do that transformation.   But what map does is   make sure the transformation works exactly the same way for all the different sources.  

[00:11:34] Unknown:

So you were mentioning how SunPy   is a   sort of Python implementation of some of the   some of the capabilities found in the IDL library. And I'm wondering if there are any issues in terms   of intellectual property or   patented algorithms that you have to   sort of steer around in the process of writing and maintaining SunPy.  

[00:11:59] Unknown:

So the ideal library isn't   licensed at all. There's no license on it. It's just   made available,   which makes it a very   sort of gray area. But it is put out there with the intention of being able to use it. But   primarily, what SunPy has done at the moment for the most part is implement   similar functionality, but in a completely different way because we wanted to implement   Pythonix   type functionality   rather than   just,   like, translating   function x from ID up to 5. Mhmm. So we've mostly avoided that by not doing it so far.  

[00:12:44] Unknown:

And you were also talking about,   particularly   in relation to the map type being able to do processing on images   from the sun. And I'm wondering if there is any computer vision routines that are implemented or available for being able to do automatic,   you know, feature extraction for, you know, certain research purposes.  

[00:13:07] Unknown:

I've done a bit of work on that personally, but none of it's ever made it into Sampi   For the kind of image manipulation stuff that we do during Sampai, which is mainly   rotate and also some kind of   transformations   to   other ones not in Sanpai yet, but being worked on is transformations   on,   to compensate for the rotation of the sun, which changes   depending on how far away you are from the equator.   So the equator rotates faster than the poles.   So if you have an image or a time series of images   and you want to,   like, rewind or fast forward the,   image, you need to do some   complex warping of the image. And all of these kind of things, we've just been using scikit image, which is an excellent Python based image processing library rather than writing stuff from scratch.  



[00:14:08] Unknown:

Are there any considerations that you need to be aware of when writing software libraries for practitioners of the hard sciences that would be different if the target audience were software engineers?  

[00:14:18] Unknown:

So pretty much my entire   programming experience has either been writing software as a scientist or for other scientists.   But I think most of the considerations   for   scientific software are in the   sort of the way it's developed.   Scientific software is pretty   unstructured in the way it's normally developed,   unless   you're talking about developing for specific   instrument, like a calibration routine for a telescope or something.   Most   software written by academics is sort of I need to solve this problem, and they'll   write something   sort of, like, basically scientific software never gets out prototyping phase.  

Because as soon as you have a working prototype, it's done. It works. It's done the job.   So it's kind of from   the building a library like SunPipe perspective.   You have to consider the fact that you have to make it general enough that everybody can use it, and it won't place restrictions on what people wanna do. But at the same time,   sort of encourage people to do things well and   use best practices,   which is 1 of the reasons why I think I think it's such a great language for scientists. Because   for a start, it forces them to indent their code, which   for a lot of computer scientists is like, well, you always indent your code. I mean, why would you not indent your code?  

I had people I've had people, especially in my undergrad, who would write code with no indentation or spaces at all. It runs fine until it doesn't, and it's really difficult to debug.   It's it's   Python itself makes an excellent foundation,   and he's really starting to gain traction within   astronomy a lot and also now slowly but surely, so does.  

[00:16:13] Unknown:

So it's really interesting. Within my organization, actually, there's been a little bit of pushback against Python recently,   because of the fact that it's a dynamically typed language. Some of our architects want us to use strongly statically typed languages instead.   But it it's kind of interesting hearing you say that Python makes kind of an ideal language for scientists because it enforces some of these best practices like good indentation and the like. Do you and do you find that lots of people writing software for the sciences   have discipline around writing tests for your code as well?  



[00:16:50] Unknown:

No. Not in general in academia. Testing,   version control,   these kind of things are still relatively rare. There's   a lot of effort,   like, excellent   projects like software carpentry   and similar kind of educational exercises,   which   are trying to teach academics   sort of software engineering skills   because   mostly academics   kind of hit   postgraduate   research. And they may have been taught how to program   like my for instance, my undergraduate programming module was about numerical methods, like solving differential equations   numerically. And,   we had an exercise where we had to   or where I he was   calculating,   like,   orbital dynamics of, like, 3 body problem kind of numerically.  

And   but not a lot of effort is   or has been at least put into teaching scientists   how   to write code well rather than just how to program.  

[00:18:02] Unknown:

So can SunPy consume data directly from telescopes and other observational apparatus?  

[00:18:09] Unknown:

I've never heard of anyone do, like, live telescope output of.   Your question did make   me wonder   it would probably be possible.   If I ever get a telescope, I might try it.   But   the kind of focus of SunPi is   science quality, like calibrated data. So   when you   put a camera at the back of a telescope,   you take an image,   say, I don't know, if we're taking pictures of, like, 30 frames a second.   Each 1 of those images is gonna be relatively   low quality,   because of   if your ground based telescope atmospheric   distortion.   You got a fair bit there to get get through before you reach the sun.   And then you've also got, like, parts of the CCD that are more sensitive than others or parts of the, like, certain wavelengths that don't filter properly. You've got a lot of calibration to do from raw camera at the back of the telescope   to data that's science quality.  

And SunPy currently doesn't do any of that,   mainly because   those kind of routines are written by the teams that design and build the telescopes   because they're the ones that can test the code,   like, calibrate it with the actual telescope and know all the ins and outs of the telescope design. So we do get   few feature requests every now and then. Like, oh, can you support calibrating this, that, or the other? We're like, we don't have the expertise or the definitely don't have the developer time to port that code or rewrite it.  

[00:19:52] Unknown:

So so, basically,   it sounds like SunPy is a   a module in that,   like you say, another team would produce the data with their telescope, and they have their own whole code base, and they would hand you the, you know, nicely calibrated,   well groomed datasets that you would then operate on with SunPy.  

[00:20:14] Unknown:

Yeah. That's the idea. And   more and more   but as the   the quantity of data gets bigger and bigger, like, for instance, NASA's most recent flagship solar fish mission,   the solar dynamics observatory,   downloads about   a terabyte 1.2 terabytes of data a day.   When that was launched, I was still doing my undergrad,   and I was looking at the specs of the satellite. And I was most stressed to realize that it could download data from space faster than I could download data on my home broadband connection.   And   so as that data gets sets get bigger and bigger, it becomes more and more impractical for people to select scientists to do that calibration themselves.   So most of the data   from   SDO   is   shipped to people   pre calibrated,   with full coordinate information in the metadata,   and that's the kind of stuff that SunPipe really works with.  



[00:21:17] Unknown:

That's that's really interesting because hearing 1 of the other podcasts in the Python space   talk Python to me, recently did an episode,   where they interviewed 1 of the researchers from the, CERN Supercollider,   and they were talking about how there was such a really substantial chunk of their software and even a hardware stack   that was dedicated to that whole process of, you know, filtering the reams and reams and reams and reams of data that these collectors would would grab   down to a level where scientists could even begin to   actively   make use of it. So it's it's very interesting that for for solar astronomy,   that's a completely different bailiwick. You folks don't even handle that.  



[00:22:02] Unknown:

It that's the way it's going. It's going more   to, like, it that being done   as a service to the scientists.   But there is   still a lot of   individual data reduction   that solar physicists do,   for older data or different sources   or ground based current ground based telescopes mostly   ship the raw data.  

[00:22:30] Unknown:

So I noticed on the project site that SunPy leverages Astropie internally. Can you describe the relationship between the 2 projects and why someone might want to use SunPy in place of or in addition to AstroPy?  

[00:22:43] Unknown:

So AstroPy is a fantastic and   substantially larger project than SunPi at the moment.   And it it's a core library for astronomy.   It provides   a lot of the functionality that SunPy relies on, whereas SunPy is a core library for solar astronomy, which is kind of a subset. So you could   do   solar physics data analysis   using just Astrappi because it provides you the tools you need to read the files,   And then you would get a NumPy array in a dictionary, and you could do your analysis.   Sunpro   provides sort of wrappers on top of that, like map I was talking about earlier,   and also more   specific solar related things like constants and   calculations for coordinate systems and stuff.  

Since I've been, like, involved with SunPower development heavily, I've been   trying to work closer with AstroPi   and sort of integrate these 2 communities a bit more.   This   is really   the most obvious example of this is probably the most recent AstroPi coordinates framework.   It was released in   ashwine 4,   I think,   which is a generic framework for holding   astrophysical   coordinates,   like points in 2 d or 3 d space,   related to some kind of origin, be it the the Earth, the sun, or something.   And   some like, myself and the Sun Pi community   contributed to the design and then the development,   the Astrobee coordinates framework.  

And   because of that, we have a nearly finished   set of solar physics coordinate frames   that use the AstroPi framework.   And because we helped   design and develop it, it ticked most of the boxes we needed. There's only a few things that we had to add or tweak, like, kind of pull requests afterwards, things we didn't think of.   And we've had there's a lot of other kind of   problems that we've solved in some pie or things we've come up against that   either AstraPie have made easier or   we've contributed   functions to AstraPie.   As of the most recent release, we   used AstraPy units everywhere throughout our code.  

Astro Pi units is potentially the best single sub module in Astro Pi   and shouldn't really be in Astro Pi at all because it's a generic   physical units package. So you take a NumPy array and you can say   my NumPy array multiplied by units dot meters.   And now your NumPy array is   a quantity with the units meters.   So if you try to   subtract   an array,   which was   quantity time from your array of quantity meters, it would go.   What on earth do you mean? You can't subtract a second from a meter? Doesn't make any sense.   And raise an error.  

Oh, like the main way we've used this in is to   simplify our API. So, like, places before we had, like, flags for   is this in physical coordinates or pixel coordinates? And now we just read the units attribute of the array that's passed in.  

[00:26:26] Unknown:

Yeah. We interviewed 1 of the core developers of Astro Pi a couple of weeks ago, and the units module, when we were talking to him, definitely sounded like a very interesting and promising piece of software that could, as you said, be broadly useful even outside of the AstroPi or the SunPi packages for anything that has to do with physical quantities.  

[00:26:47] Unknown:

That's cool. Who did you interview?  

[00:26:49] Unknown:

Eric and his last name is escaping me.  

[00:26:52] Unknown:

Toru? Yes. Toru. Yes. Oh, very   nice. He's the guy that wrote   or leaded the coordinates.   The other thing that we've   sort of that's come out of,   me being   involved in the astropical   community is the Python and astronomy conference, which is   brought together different people from different   fields and projects,   but all people who are working with   Python and doing astronomy or astrophysics. Mhmm.   And we had a really good conference last year in April, and there's another 1 coming up in March, And they've been really constructive of that 1.  

[00:27:35] Unknown:

And so looking at the documentation,   I got the impression that there's a fair amount of visual representation of the data for analysis.   Can you describe some of the challenges that that is posed? And is there any integrated support for project Jupyter or any other graphical environments that SunPy supports?  

[00:27:52] Unknown:

So most of   the SunPy visualization   routines   have used map.   So   if you create a sun pi plot in a Jupyter notebook, it behaves as you would expect it to.   1 of the   more recent changes to some plotting   is   the   fantastic   WCS axis package,   which   has been   the the development of it's being led by Tom Robitaille, who's 1 of the core Astro Pi devs.   And WCS access is a package that allows you to do,   like, kind of 2 d image plots, but other 2 d plots in that plot   and have   your   coordinate axes,   your coordinate frame,   and your axes not aligned with the array. So   you could imagine having   your   square   array   and your physical coordinate system rotated 45 degrees.  

And when you create a plot with WCS axis, all of the, like, image coordinate grids and the ticks around the edge of the frames would all be aligned to the physical axis. It's like 45 degrees to the,   around the edge on the edge of the map lock the plot   and allows you to do, like, mouse over and overlays and stuff in different transformations.   And that's made a big difference to   map visualization   routines.   The other   the only other kind of visualization   stuff that Sunpies had some involvement with is the Genga project, which is a   graphical tool for viewing physics files. And, we've had done a bit of work to support   improve the solar physics support in ginger.  



[00:29:50] Unknown:

So what are some of the most interesting applications that SunPy has been used for?  

[00:29:56] Unknown:

I don't have a a lot of   specifics. I mean, I know a few,   research projects   that have used some pie.   1 of my colleagues has done   some work studying waves in the   solar chromosphere,   and he used some Python Python for a lot of that.   Another 1 of my,   friends has done,   built a map,   like, kind   of graphical tool where he could   click on the images to measure how   features   of the solar limb change with time   and record those.   I think 1 of the most interesting,   albeit currently hypothetical uses for SunPi   is the new,   telescope   that's being built on Hawaii.  

It's gonna be the   by quite a margin, the largest solar physics telescope that's ever built ever been built. And we're currently   discussing with that project about how they're gonna use Python and hopefully also some pie.  

[00:31:16] Unknown:

And so before we move to the picks, are there any other questions that we should have asked but didn't? Or anything else that you'd like to bring up?  

[00:31:24] Unknown:

Not that I can think of.  

[00:31:26] Unknown:

Alright. With that, we will move to the picks. And for my first pick today, I'm going to pick the Elm language,   which is a functional language that compiles to JavaScript.   And 1 of the things that it says on the   homepage for it is that if you can successfully compile it, then it's guaranteed to not have any runtime exceptions, which is   quite a bold claim, and everything I've been hearing about it is very positive. And I'm actually planning on using it for a project that I'm going to be starting fairly soon. So definitely recommend checking that out.   It just seems like a really   interesting and useful way to do front end development.  

My next pick is the Avro serialization   specification.   And what makes this interesting is that it is a binary format,   and it also   keeps the schema of the data   collocated with the data itself. So the actual   serialized   form of the message includes the data schema. So you can actually   change the schema of your data over time   but you will still always be able to read the data that's been serialized because of that   collocation,   the locality of the data.   So I've   been considering using that for another upcoming project that I'm going to be doing soon at work.   Definitely seems like 1 of the most   flexible serialization   schemas that's available in terms of being able to actually   maintain the   semantics of the data along with the data itself.  

And for my last pick, I'm going to choose a website called Common Sense Media,   and it is a site that will give   you useful reviews and summaries   of various media, whether it's movies, video games, books.   And it's a good way for parents to be able to   check up on a movie or game that they're considering letting their kids watch or play without having to actually   watch the entire thing beforehand and just get a good overview of what's the level of violence or profanity and  

[00:33:42] Unknown:

Thanks, Tobias. My first pick is a website. It is called Massdrop.   And,   I will say that if you are in any of these sort of groups that Massdrop panders to,   this thing is dangerous.   It the the basic idea is that   they will have   agreed upon   drops.   You know, there are different subject areas, like,   there's 1 for keyboards. There's 1 for keyboards and input. There's 1 for personal carry, so it can be anything from   wallets to small tools, or whatever the case may be. There's 1 for hobbies, so it has things like   drones,   or, like, they just recently sold a set of lock picks and a practice lock, so you could learn lock picking.  

It's basically   and the and the idea is that   if you get in on a drop,   then the more people who join the drop, the lower the price of the item becomes,   to within a certain threshold.   And it's just it's it's a really neat way to, a, save some money on some interesting things, and, b, find out about some interesting things that you may never have heard about. I the the wallet that I bought, my Ridge Wallet, I I I found out about and bought through,   Massdrop that I picked in the previous episode. It it's just it's a great site.   Someone who people have pointed out that you don't always necessarily save a ton of money through Massdrop, but,   you know, it's it's it's never   you never end up overpaying, I'll say that much, and I have definitely provably found instances where I've saved, you know, a fair bit by using it.  

My next pick is,   a beer called 21st Amendment Fireside Chat.   This is an interesting 1. It is   a a   spiced English ale, and it is a winter beer.   Very smooth, very tasty.   It's just a really good solid   winter ale with very little hop signature, nice and malty.   And my last pick is a YouTube show.   This kind of an interesting sort of mishmash. It's called Extra Credits,   and it was originally   meant to discuss   games and the games industry   video games, rather, the video game industry and video game design.   But,   through this sort of sideways,   I don't wanna say accident isn't quite the word, they ended up going in-depth about this 1 particular game that was set in a particular period in history,   and they had such huge crowd,   response to that that they now have this side series   where they go through and discuss various   events and periods in history,   nothing to do with gaming. It's kind of like so it's like game video games, video game design, and history. It's an interesting show, definitely worth a look.  

So, Stuart, what kind of picks do you have for us?  

[00:36:41] Unknown:

I have a few.   A few weeks ago, I think 1 of you picked out you are listening to as the background noise thing. That would be me. Love it. Yes.   I looked well on that and went, oh, that's interesting, but it needs more space.   So I discovered that there is   a live stream from the International Space Station,   that includes the ground to earth radio   as an audio track.   So for the last few weeks, I've had that running in the background.   I also   little after that discovered that there is a   high definition video feed from some cameras that was recently attached to the outside of the International Space Station,   and that streams live   7 20 p HD video of the earth, which is just beautiful to have on in the background.  

My next pick is the YT project, which I thought I would   pick out specifically.   It's a Python project that I used extensively during my PhD research, which has been mostly numerics based.   And it   has   abilities of reading   sort of, numerical datasets and do   visualizations,   specifically   like   slices   through 3 d domains.   The yt slice bot interface is 1 of the   nicest bits of Python code I've ever had the pleasure to use.   And,   thirdly,   a project that I discovered   this week is called half   half studio gear, which is   a set of live   audio processing   plug ins,   that run on the jack audio server. I think it's Linux only,   but it allows me   to process the sound coming out of my microphone   with things like an equalizer or in a gate and a compressor,   pipe them back out to   false audio and then through to Skype or Mumble or any of the other   places I want to send.  



[00:38:55] Unknown:

Very cool. For what it's worth, I don't know about CAF because I've never I've never actually used this particular thing,   but I know for a fact that Jack runs on OSX as well.  

[00:39:05] Unknown:

Yeah. Jack does. I don't know about Kaf. I think   I think it's I think Kaff is a Linda Coney, whereas Jack isn't. But Jack is also awesome.  

[00:39:14] Unknown:

Gotcha.  

[00:39:16] Unknown:

Alright. Well, we really appreciate you taking the time to join us and tell us all about SunPy. For anybody who wants to keep in touch with you and follow what you're up to in the SunPy project, what would be the best way for them to do that?  

[00:39:29] Unknown:

Well, I have   Twitter. My personal Twitter is at Stuart Munford,   and the project is at sunpie project.   There is also the Sunpy website, sunpie.org.   The GitHub page, github.com/sunpy,   and also our   IRC channel on the free node dash sunpy,   which is a good place to come if you wanna have a chat with me.  

[00:39:57] Unknown:

Great. Well, we really appreciate you taking the time out of your evening to join us, And I'm sure that our audience will enjoy learning more about solar physics and Sun Pi.   Thank you very much. Thank you. Have a good evening. Thank you.