Research Platform Services Blog
Month
By now you might have noticed that here at ResBaz we have a thing for open data. And so had you during ResBaz 2015!
We believe that all data (and code) should be open to the public.
Philippa Karoly (who some of you might know as one of our excellent MATLAB helpers) and Ewan Nurse recently wrote such a paper. They went through the extra effort of sharing their collected data through GitHub and figshare. In their (soon to be published) paper, they analyse EEG data using MATLAB, producing some meaningful and beautiful images of brain activity. Data that could potentially be used to control quadcopter with your mind!
And with Philippa’s and Ewan’s help, we created a short MATLAB lesson showcasing their stunning results.
The lesson even made it into the official Software Carpentry curriculum and can now be accessed and taught from all over the world.
How about you? Do you have interesting data that you would love to share with us? Maybe you are even thinking about using one of the research tools we teach to analyse your data? Come chat with us at Hacky Hour!
Recently we taught a Beginner’s MATLAB workshop for researchers. The workshop was a definite success, with 25 researchers coming along despite the first day unfortunately falling on a public holiday.
Throughout the week we covered a lot of material, including the command line, scripting, for-loops, conditionals and functions. On the last day we ran a quick refresher on all the topics we’d covered so far, and also discussed good programming habits, error handling, and on-line resources.Despite public holiday, @MATLAB workshop is well on the way. That’s #dedication! #resbaz @ResPlat @EngKerry pic.twitter.com/U1QefoLr61
— Isa Kiko (@Isa_Kiko)
June 8, 2015
During the course of the week I had the opportunity to sit back a watch the very talented Isabell (@Isa_Kiko) teach the class about conditionals using the metaphor of choices we make in our everyday lives. These choices might be anything from choosing whether to take an umbrella (which is dependent on whether you are going outside, and whether it’s raining), to choosing whether you should eat (which happens if you are bored or if you are hungry). I was in a bit of a contemplative mood at the time, and these metaphors got me thinking about how programming can be applied in everyday life. Of course in the present day we see examples of logical programming cropping up everywhere, even for the most useless of purposes.Day 2 of the @ResPlat @MATLAB: @Isa_Kiko rocking scripts and loops pic.twitter.com/IgW7vxH7MD
— Kerry Halupka (@EngKerry)
June 9, 2015
“There is no more worthy purpose of programming than confusing cats” - Bill Gates (maybe)
However, sometimes I want more. Anyone who has ever made a mistake and wished there was an undo button for life, or a previous save point you could just jump back to, understands what I mean when I say that some features of programming could be really useful for life. I came up with a list of ways I wish life was more like MATLAB, let me know what you think should be added to the list!
1) Commenting
We teach our workshop attendees the importance of commenting their code. Anyone who has gone back to a script they wrote months ago will understand the importance of commenting. Isabell explained it by saying “If I come back to my script in 5 months, I want to know what the hell I was doing there, so I comment”. I can’t remember the number of times I’ve cursed past-Kerry for not commenting a line that she thought was so clever at the time. Sometimes I wish I could add comments to bits of my life, so that when I look back on something I did months ago, I don’t automatically assume I was out of my mind because the reasoning seems insane. However, I assume that a lot of the comments would be incredibly unhelpful (ie. “I did this just for lols”) and make me wish I had some kind of time machine to go back and fix them. Which brings me to the next item on the wish list…
2) Debugging
While it may be good programming practice to plan ahead to avoid errors in you code, or fix errors prior to execution, we don’t always get it right. So MATLAB is kind and provides a lot of different debugging aids. I wish life had this feature. Wouldn’t it be great if life placed a little red squiggle below errors we’re about to make, and suggestions for avoiding them? For instance, “You might not want to eat another deep-fried mars bar”, “Maybe you shouldn’t pull an all-nighter” and “Instead of watching funny animal videos, you should probably do some work”.
3) Breakpoints
Imagine life with a pause button.
Go on, do it.
If you’re like me, there’s probably been at least one time today that you would have appreciated the time to stop and think before you did something. For me, it was right before the third coffee. Yesterday, it would have been just before pressing “Continue Playing” on Netflix at 2am. MATLAB does have that capability, because sometimes things can go wrong while programming , and the reasons why are not always obvious. So being able to stop your program just prior to where things start to go wrong, and stepping through line by line to see the values of the variables is a really useful asset.
4) Parallel Computing
MATLAB has the ability to execute loop iterations in parallel, so that all of the repetitions of the loop are run at the same time. It’s pretty clear how useful this would be in life. One (hyphenated) word: Multi-tasking. I can see it now:
5)
Logic
If all of the decisions that
everyone made in life were based solely on logic, it would be trivial to
understand someone’s motives for any action. We would never have difficult
decisions to make because our truth tables and logic gates would have
instantaneously arrived at the most correct answer. Case in point, me at a café
5 hours ago:
Should I order caffeinated or decaffeinated coffee?
Logical Me: It’s past 6pm, I won’t be able to sleep until at least 2am. Better
not get the caffeine.
Illogical Me: ALL THE CAFFEINEZZZ
When I started writing this post, this logic driven, commented and debugged life sounded idyllic. But without my irrational decision making and poor life choices, my sleep-deprived, caffeine-addled brain would not have produced this article. I’m not sure whether that is an argument for or against a life more like MATLAB, maybe in the end it’s best that we as humans keep doing what we’re good at, and let the programs be the logical, sensible ones. What do you think? Should life be more like a program? And is there any software package that has better life features than MATLAB? Share your thoughts with us by tweeting us at @ResPlat or @EngKerry.
By: Katie Ewing
After organizing the ResBaz Conference, held at the University of Melbourne in February 2015, you could say I was having a bit of conference withdrawal. No more new connections. No more free food. No more fun! Luckily, an opportunity arose to present about ResBaz at the Universitas 21 Graduate Research Conference in Shanghai, China. How could I resist?
I’m back! Which way for ResBaz?
Founded in 1997, Universitas 21, or U21 for short, is the leading global network of research universities for the 21st century. Their vision is to “work together to foster global citizenship and institutional innovation through research‐inspired teaching and learning, student mobility, connecting our students and staff, and wider advocacy for internationalisation” [1].
Right now, there are 26 member universities around the world. That’s a whopping 1.3 million students and 210,000 academics and researchers part of this network! [2]
Hmm…which U21 uni will join the Research Bazaar next?!
U21 holds an annual Graduate Research Conference (GRC) to give graduate students an opportunity to present their research and network with other students outside their fields. The theme of the conference varies year-to-year, but draws students from a broad range of disciplines.
The 2015 GRC was held at Shanghai Jiao Tong University (SJTU) in Shanghai, China from June 9-12. Sound familiar? That’s because SJTU were the original publishers of the Academic Ranking of World Universities each year (where Melbourne Uni proudly sits at No.44). SJTU was founded in 1896 and is one of China’s oldest and most prestigious universities. Hosted at the School of Media and Design, the conference’s theme was “Digital Future,” with the aim of exploring the evolution and influence of digital technology on society.
SJTU’s School of Media and Design focuses on the interconnection of technology, science, research and arts. The perfect setting for all topics Digital.
Each member university could nominate graduate students from different departments to present their research covering multiple disciplines all related to the digital age, from digital media to digital life to digital technology. Representing Melbourne Uni, we had Prasanna Sritharan (Mechanical Engineering), Voula Karvelas (Education), Suneel Jethani (Media and Communications), and myself. I was fortunate in that I didn’t have to present my own research, but instead could talk about ResBaz!
The Melbourne Uni contingent (L to R): Pras, Katie, Voula, and Suneel.
The conference included 3 days of keynote talks from leading SJTU researchers, student presentations, a digital company tour, and of course nights out in Shanghai! Stay tuned for my next post, where I’ll fill you in on all the juicy details of the actual conference itself!
[1] About Universitas 21: http://www.universitas21.com/about
[2] If you’re reading this and are doing or thinking about starting a PhD at an U21 university, you may want to consider a jointly-awarded PhD program with two U21 universities. Joint PhD projects create opportunities for research collaborations with high quality research-intensive universities around the globe. You can truly tailor-make a PhD program for yourself!
1. “Join the Shark’s Den Challenge and start creating now!” by Aliza Wajih
Aliza launches our new 3D printing initiative, an innovative Shark’s Den challenge happening throughout July! We still have some spots open so click the link above & check it out.
2. “Meet Louise - 3D Modelling Medical Scans!” by Louise van der Werff
Louise is one of our newest Research Community Coordinators looking at the potential of 3D modelling using medical scan data!
3. “MATLAB – A Love Story” by Kerry Halupka
Kerry is also new to us. Her love and passion for MATLAB makes her an excellent Research Community Coordinator!
Hey there,
My name is Kerry Halupka, I’m a PhD student in Biomedical Engineering, and I’m the newest member of the Research Platforms team. Since joining the team I’ve been involved with teaching free MATLAB workshops to researchers, and tweaking the course materials.
I had a ball teaching the @MATLAB for Women workshop with @Isa_Kiko last week at @unimelb. Looking forward to more in the future! @ResPlat
What did our attendees think of our women-only @MATLAB workshop at @unimelb? Find out here! http://t.co/1yp72cvRh1 #ResBaz
— Research Platforms (@ResPlat)
May 18, 2015
I started my PhD at the University of Melbourne in March 2013 after completing a Mechatronic Engineering degree at The University of Adelaide. My current work focuses on the bionic eye that is being developed by Bionic Vision Australia. By recording activity in the brain while activating the retina with our retinal prosthesis, I’ve been developing a computation model built in MATLAB to predict the brain’s responses to stimulation. My aim is to use the model to optimise the way that we electrically activate the retina, hopefully increasing the resolution of the images our patients can currently see.
I’ve been using MATLAB since the first year of my bachelor’s degree. Way back when I thought that university couldn’t possibly be any more difficult than year 12 in high school. I realised how very wrong I was at about the same time I discovered that a steady diet of vending-machine items and the occasional all-nighter does not a healthy person make.
MATLAB was introduced briefly as a side aspect to one of my courses. At the time I was taking another course in which the programming language C was the main focus, and I distinctly remember wondering why anyone would bother with the intricacies of C (pointers were the bane of my existence at the time) when they could instead languish in the supporting arms of MATLAB.
Since that first course MATLAB and I have had a lot of time to bond. I’ve spent many a late night bathed in the light of my computer screen as I type “just one more line”. MATLAB has continued to enchant me with it’s beautiful development environment; the helpful, yet not pushy error messages (even going so far as to quietly hint at errors before I even execute my code); and the huge and incredibly helpful community of people around the world sharing their own functions (I often say that for anything you want to do in MATLAB, there is very likely someone who has already done it and shared their code, or there is a blog/discussion thread surrounding the issue). I’ve lost count of the number of times I’ve stumbled upon a new, exciting function and just think “I can’t wait till I have a chance to use that!” (this is my current obsession).
Through the years MATLAB and I have had a complicated relationship. I spiced things up by introducing MATLAB’s graphical programming environment Simulink as a third party during my honours year. Together we made some lasting memories – I used Simulink to build a biomimetic detection and pursuit algorithm modelled on specific neurons in the visual pathways of certain insects, and tested the model in a virtual arena.
(a) Virtual target pursuit arena built in Simulink (b) Camera viewpoint 1 (Cam. 1) records the chase from above. © Image information from the pursuer viewpoint (Cam. 2) undergoes analysis through algorithm in Simulink.
I had a brief fling
with Python during the first year of my PhD while undertaking a computational
neuroscience course in Okinawa, Japan. Suffice it to say that strained our relationship.
However, I maintain that it was nothing but a meaningless holiday romance. Between
you and me though, I’m still a little bit intrigued by Python, there is beauty in imperfection after all, I sometimes
consider taking one of the Python courses taught by Research Platforms.
In my other life (you know, the one that doesn’t involve a thesis) I enjoy martial arts, cooking, eating and music. Sometimes my love of eating and MATLAB even cross over!
Love it! @Msmiciobiocode made an animated pie chart showing how much pie has been eaten! @MATLAB @EngKerry @ResPlat pic.twitter.com/0aviPZ3F8m
— Isa Kiko (@Isa_Kiko)
June 9, 2015
I’ve always loved learning new things - I’m half way through about 3 different courses on Codecademy right now, and I’m also learning German from another of the Research Community Coordinators, Isabell. So I’m a big proponent for learning. Since becoming a Research Community Coordinator I’ve realised that teaching is just as enjoyable and rewarding, if not more so. The two sides of this educational coin are also complimentary - every time I’ve taught MATLAB so far I’ve learnt something new myself. So if you think you might want to learn along with me, come to one of our MATLAB courses, or email me at kerry.halupka@gmail.com or tweet at me (@EngKerry) for more information (I promise I won’t wax lyrical at you about my love for MATLAB).
By Damien Irving and David Flanders
The second annual Mozilla Science Lab Global Sprint (mozsprint) has just wrapped up (see the global summary here), so we thought we’d recap the action here in Melbourne:
Still going strong here at #mozsprint #Melbourne, some break out groups doing training as well. pic.twitter.com/vHkFagq8X3
— David F. Flanders (@dfflanders)
June 4, 2015
The reinvigorating effect of #mozsprint in one picture #50ShadesOfGreen @ResPlat @MozillaScience pic.twitter.com/e65U6GBQXX
— Isa Kiko (@Isa_Kiko)
June 12, 2015
Melbourne team on our last hours of #mozsprint! Balloons, cake, wine - oh my! @dfflanders @Isa_Kiko @cobismith pic.twitter.com/EAiNT3toyR
— Dejan (@heyDejan)
June 5, 2015
This idea of getting our Research Community Coordinators and software developers together in the same space for two days of hacking proved so successful that we are going to run our own internal “ResHack” event every three months (with every fourth coinciding with the annual mozsprint). Thanks again to the team at the Mozilla Science Lab for hosting such a great event!
By Damien Irving.
This time last year we introduced you to the team of geophysicists behind Underworld, which is a computer model that simulates the movement of continents (i.e. plate tectonics). Every year Masters students in the School of Earth Sciences are required to take a number of one-week courses offered by the Victorian Institute of Earth and Planetary Sciences (VIEPS), so Professor Moresi and his team offered an Underworld VIEPS course for the first time in 2014. Their model has a Python front-end, so we offered to help out by delivering a half-day introduction to Python at the beginning of the week.
These geologists are rocking the @swcarpentry shell lessons! #punintended @ResPlat pic.twitter.com/nGvdyoINfv
— Damien Irving (@DrClimate)
May 25, 2015
Upon reviewing the 2014 course, Professor Moresi found that in order to use the Underworld model more effectively, his students would require a stronger grounding in the fundamentals of Python programming. To address this issue, an “Introduction to Python” course was added to the VIEPS schedule for 2015. We helped out by running the first two days of that course as a regular Software Carpentry workshop, and then for the remainder of the week the students applied their newly acquired Software Carpentry skills to typical Earth Sciences problems (e.g. analysing and plotting seismic data). This introductory course was set as a prerequisite for the Underworld VIEPS course, which was held two weeks later.
Recipe: Take @EarthSciMelb Masters students, add @swcarpentry + lessons on #iris & #cartopy, get awesome #dataviz pic.twitter.com/s0r09TQEN1
— Damien Irving (@DrClimate)
June 16, 2015
In the lead up to the Underworld VIEPS course, we also worked with Professor Moresi and his team to make the Underworld model available on the NeCTAR Research Cloud via resbaz.cloud.edu.au (which is powered by our DIT4C engine). This ensured a seamless student experience with zero software installation issues, which made life much easier for the workshop instructors. We are now in discussions with other researcher groups about improving the accessibility of their software via the NeCTAR Research Cloud / resbaz.cloud.edu.au, so please do get in touch if you think this approach might work for you!
Written by David F. Flanders during #MozSprint.
As Research Community Manager for the University of Melbourne, the most frequent conversation I have with researchers goes something like this:
Researcher: “Hey Flanders, thanks for the training your team did on <ResearcherToolXYZ> the other day; but have you heard of this new tool called <ZYX>?
Flanders: [Loading the tool on my tablet ad hoc] “Oh wow, this new researcher tool looks brilliant, how are you using it for your research?!”
Researcher: “I’d love to use the tool but I need help to get it working so my research team can collaborate in using it together.”
Flanders: “You’ve come to the right place, let’s figure out a way to bring like-minded researchers like yourself together to start using this tool collaboratively; and I’ll talk to our dev team to get it up on our Research Cloud!”
I’ve had the above conversation over and over again, and it is only increasing in frequency. Though don’t take my word for it! - Check out this growing list of Research tools on Wikipedia (via our DVCR, thanks Liz!)!
In addition, we ran a survey at the beginning of the year (as part of our “Research Bazaar” conference) where we surveyed over 400 researchers througout Australia (early carreer, average age 33) asking them “what next generation digital research skills (aka tools, apps, datasets) are you working with right now?”
Below, over 200+ posters by researchers (ECRs) listing the current tools they are using for their research:
The most popular “next gen digi res skills” training at ResBaz:
Having worked in the sector for over 15 years (accross three contients and dozens of instituions) I was blown away by this researcher response, where we not only saw well over 250+ tools mentioned but also saw a massive crying out for more opportunites to work in cross disciplinary environments.
[Go on, step out of your castle/cathedral on campus and visit us in a common marketplace of next gen digi res skills - welcome to the bazaar!]
Stay tuned to the ResBaz News Channel (aka Twitter hashtag) for more tactical thinking on how we can support researchers as they use more and more data intensive research tools in the cloud.
Preview of future post: We run an OpenStack Cloud which allows us to do the heaving lifting of new research tools+data via our DevOpps team! With the tools installed on our cloud we then bring researchers together in collaroative team training situations so new skills can be shared and new friends (co-authors) can be made.