My poster at Dynamics Days 2014

Presenting this poster at Dynamics Days 2014 today: “Uncertainty in the bifurcation diagram of cardiac action potential duration ” or “Polynomial chaos uncertainty quantification of a return-map model of cardiac action potential duration restitution.”

Here’s a JPG of it for anyone interested: JPG version of poster presented at Dynamics Days 2014 by Caroline Ring

 

And if you really want it, a large PDF version: Ring_DD2014_poster_final

Visualization with color

Here are some neat resources I came across recently about using color in data visualization:

Why Should Engineers and Scientists Be Worried About Color?, by Bernice E. Rogowitz and Lloyd A. Treinish. This one goes back to 1998 — but it’s still relevant. Shows loads of fantastic examples of how our perception of color makes a difference in what we get out of a visualization.

ColorBrewer 2 — a trove of useful color schemes for presenting data of various sorts. It’s focused on map data (cartography) — but the schemes are just as useful when visualizing other kinds of data! You specify your data type and number of data classes, and can request color schemes that are colorblind-safe, print-friendly, and/or photocopy-able.

For general colorblind-safe checking, I like Color Oracle. It’s a free application that temporarily adjusts your screen colors to simulate what people with three different kinds of colorblindness might see. Available for Linux, Windows, and Mac.

Any other cool color visualization resources to recommend?

 

Dowsing for hepatitis

The Guardian reports that Egyptian liver specialists say they’ve developed a device that can remotely scan patients for hepatitis C. It’s called C-FAST, and it’s based on a “bomb detector” used by the Egyptian military (along with many other countries). (The guy who sold that “bomb detector” has been charged with fraud.)

Essentially, it’s a car radio antenna attached to a box, and it’s supposed to swing like a compass needle towards people who have Hep C, while staying still in the presence of people who don’t. Much like a dowsing rod. According to its inventors, it responds to electromagnetic signalling from the Hep C virus, which according to them, “vibrates” at a certain frequency.

C-FAST’s inventors claim that “it’s been successfully trialled in 1,600 cases across three countries, without ever returning a false negative result.” I note a distinct lack of the terms “double-blind” or “controlled” here.

The only publication I can find is this 2011 EASL poster abstract describing a study of 879 patients at a hepatology clinic in Mansoura, Egypt. The results from C-FAST were compared to the results of RT-PCR screening. There’s no information at all about whether the PCR screening was done before or after C-FAST screening, nor whether the study was double-blind.

The Guardian article in the first link quotes Dr. Saeed Hamid, of the Pakistan Society for the Study of Liver Diseases, claiming that he has tested C-FAST “in a blinded fashion,” and it was successful. However, I can’t find any publications at all about the study or studies he’s referring to.

I’m extremely skeptical of the proposed mechanism of C-FAST (in case you couldn’t tell). And without a lot more peer-reviewed published data, from double-blind, controlled studies, I’ll remain extremely skeptical of the amazing sensitivity and specificity claimed by its inventors.

Pinterest: first impressions

No one on the internet will shut up about Pinterest. So I signed up to see what all the fuss was about. Right now, I have a few user experience suggestions.

First of all, Pinterest, why the heck are you so insistent about requiring me to sign up through Facebook, especially if you immediately ask me to create a separate Pinterest user name and password? I’d really rather not have things linked to Facebook by default. The only benefit of Facebook sign-in, in fact, is that I don’t have to create yet another user name and password. But apparently not here. You just really super extra want to scrape my Facebook friends list to auto-populate my Pinterest friends list? Is that it?

And, okay Pinterest, I understand suggesting people to follow based on my interests. But did you really have to automatically sign me up to follow ALL of the pinboards from each of those people? Because not everything they post is actually related to the interests we have in common. When I put “design” as an interest, that means I’m interested in following pinboards with actual design stuff. It doesn’t mean I’m interested in following a Stupid Lolcats pinboard just because it was created by someone who also has design pinboards.

If you’re going to take my listed interests and automatically have me follow stuff tagged with those interests, please apply those interest tags to pinboards, not users.

Yes, on other social networks, you follow a given user, i.e. you get a feed of all of the stuff they post. When I follow @BadAstronomer on Twitter, I can’t choose to see just his posts about astronomy — I also see his posts about politics and religion. Now, I’m okay with that in the case of @BadAstronomer. But I have unfollowed people on Twitter because while they occasionally tweeted about stuff I was interested in, they also tweeted a lot about stuff I’m not into. It was a signal-to-noise problem. When there was too much noise, I had no choice but to unfollow the whole user — there was no way to filter.

Pinterest has a built-in filter, by letting you have different pinboards for each topic and letting people only follow specific pinboards. So why on earth not use that filter by default?

Also, is there any way to manage my friends list and which boards of theirs I’m following? Because right now I’m doing it by finding a post I’m not interested in, clicking on the poster’s name, and then clicking “unfollow all” and then “follow” on whatever specific pinboards of theirs I actually care to follow. Tedious.

And since I was started off following a bunch of people and boards I may or may not actually be interested in, I have to do a lot of this.

Right now, my feed is about 90% stuff I’m not interested in. And I am bored and annoyed with wading through it to set up filters to catch the cool stuff.

Overall, probably not the “Oh wow cool!” addictive first experience Pinterest was going for.

Head First Labs: workbooks for smart n00bs

Web development n00bs (like me): I highly recommend the Head First Labs books.

For a review, I’m working through Head First HTML with CSS & XHTML. I’ve learned HTML and CSS by the seat of my pants, in bits and pieces — very much on a “need to know basis.” So while pretty much everything in the book is review for me, it’s still good to have it actually organized for once. And I am finding a few gaps in my knowledge, which I’m filling in.

So far, they really seem to have nailed the style for “workbooks smart people can use to teach themselves stuff they don’t already know.”

The authors clearly did a lot of research into how people learn, and then designed this presentation: conversational, engaging, heavy on the examples and experiments. (Plus self-aware corny, silly humor — which is a great stealth teaching method. I may groan, but I’m going to remember the information in that joke.)

They’re explicit about what knowledge they do and don’t assume. This is key. I cannot tell you how many times I’ve been trying to learn something, run across some unfamiliar concept or vocabulary tossed off without explanation, and ended up spending frustrating hours trying to go back and fill in the blanks.

In fact, I just did this. Trying to learn methods for propagating uncertainty in my ion-channel model by consulting an expert and reading the textbook he recommended, only to wind up spending weeks digging up more basic texts on probability theory so I could understand the abstract applied-math presentation of the method — all of which got me no closer to actually being able to do anything with uncertainty in my model.

You know when I finally started to figure it out? When I stopped banging my head against the textbook and picked up a shorter paper with a specific example of the method being used.

Examples — that’s what I thrive on when I’m learning something new. I build up theory from observations. It’s hard for me to go the other way when I’m learning something new.

That’s another thing the Head First books are great at. You’re making stuff that works from the very first chapter. You’re constantly encouraged to experiment, try stuff and see what happens. I do that anyway — but it’s awesome to have it be part of the presentation the whole time.

As soon as I’m done with the review in the first book, I’m going to move on to Head First HTML5 Programming — which I am super excited about. (It contains an intro to Javascript, though there is a separate Javascript book too.)

I’m always happiest when I’m starting a new project and learning something new. Getting to teach myself HTML5, CSS3, and Javascript? From sources really well-designed for me to do that? So excited right now.

(Head First has books on all kinds of tech topics, not just web development. Check out the whole books list.)

Getting unstuck by writing longhand

In my last post about stuckness, I said that talking to other people is a good way for me to get unstuck. But obviously, that’s not always possible. Most of the time, with my Ph.D work, I’m working by myself. So how do I get unstuck when I can’t bounce ideas off anyone?

I do it by writing.

I write my thought process down in real time. I have many pages of notes that look like this:

[pasted plot] It really looks like a sigmoidal function of voltage. So how do I model this?

Need to use Boltzmann equation

[some math follows]

Wait, that can’t be right, it can’t activate at that voltage.

[more math, correcting the error]

OK, that’s right. But what about […and so on]

Keeping this kind of stream-of-consciousness journal works a lot like bouncing ideas off someone else.

It helps me stay on track. My entire train of thought is committed to paper. I can see what I’ve already considered and why I discarded it. I can also see exactly where my questions are, and what errors I already caught.

It also helps me unstick myself by forcing me to put into words (sentences, paragraphs) exactly what I don’t understand. Once I’ve done that, it’s much easier to identify the pieces of information I need, and figure out where I can go to get them.

And just the act of writing longhand seems to help me get in and stay in a state of flow. When I did a journaling workshop with Carol Henderson, this phenomenon of longhand flow came up. It works when you’re journaling for yourself, or doing creative writing. But it works just as well when you’re doing science.

So I keep legal pads and pens nearby. My absolute favorite legal pad is this Ampad one: college ruled, three-hole punched, with a nice thick cardboard backing so you can write anywhere. And my current favorite pens are these Paper Mate ballpoint pens. Yes, I like them in purple. (No, Ampad and Paper Mate aren’t paying me anything or giving me anything. I just happen to like those particular products.)

I picked up both of these office supply habits from one of the engineers who taught me the most when I started graduate school, Ned Rouze. Ned’s notes are always tidy and orderly and clean; mine aren’t. But he taught me the necessity of always keeping a legal pad and pen within grabbing distance.

So when I’m stuck, I can either bounce ideas off someone else — or I can bounce them off a yellow legal pad with a purple pen. Either way, it helps me get untangled and back into the flow.

Royal Society historical journal archive now free, open to public

The Royal Society has been publishing peer-reviewed scientific research since 1665. Now, they’ve thrown open the doors to their historical archive. Everything more than 70 years old is freely available.

It’s science history gold. As the Royal Society’s announcement notes, you can now get a copy of the following articles:

Franklin, B. “A Letter of Benjamin Franklin, Esq; to Mr. Peter Collinson, F.R.S. concerning an electrical Kite.” Philosophical Transactions 47, 566-567, 1751-1752. Free full text.

Newton, I. “A Letter of Mr. Isaac Newton, Professor of the Mathematicks in the University of Cambridge, Containing His New Theory about Light and Colors: Sent by the Author to the Publisher from Cambridge, Febr. 6 1671/72; In Order to be Communicated to the R. Society.” Philosophical Transactions 6, 3075-3087, 1671. Free full text.

The latter has a wonderful abstract (given in the issue’s table of contents), which I here reproduce:

A Letter of Mr. Iſaac Newton, Mathematick Profeſſor in the Univerſity of Cambridge; containing his New Theory about Light and Colors: Where Light is declared to be not Similar or Homogeneal, but conſiſting of difform rays, ſome of which are more refrangible than others: And Colors are affirm’d to be not Qualifications of Light, deriv’d from Refractions of natural Bodies, (as ’tis generally believed;) but Original and Connate properties, which in divers rays are divers: Where ſeveral Obſervations and Experiments are alledged to prove the ſaid Theory.

I think all abstracts should be written like this, particularly including the long esses. (I can’t help mentally pronouncing them as f’s, and then the whole thing takes on a lovely Mel Brooks sort of humor.) It would make literature search so much more fun.

(However, I’m glad that my papers need not include the sentence “Amidſt theſe thoughts I was forced from Cambridge by the intervening Plague…” Streptomycin is a wonderful thing.)

On stuckness and teamwork

From a discussion thread on Making Light today, a link to JeffWofford.com:

“A programmer’s greatest enemy is getting stuck.”

Go and read that post. And when you’re done, read the following posts, “Four Kinds of Stuck” and “Keep Taking Steps Forward.”

I recognize many of the specific examples from the days when I was getting paid to develop software, and from the programming I do now in the course of my Ph.D research. I’m mentally applying the concepts to the bigger picture of my Ph.D, though.

Doing a Ph.D. is, by its very nature, an exercise in being stuck somewhere between Stuck in the Muck (knowing what to do, but not knowing how) and Conceptually Stuck (not knowing what to do). It wouldn’t be original research if you didn’t have to figure out how to do something that hadn’t been done before.

The problem is, at the same time, you’re learning how to properly do the things that have been done before — and you’re still making mistakes on some of those things. So when something isn’t working, it can be hard to tell whether it’s an easy problem you’re just screwing up, or whether it’s a hard problem you’re properly working through. That makes it hard to tell whether you should say “Okay, this isn’t the solution, I’ll figure out a different approach” or “I have to keep working on this until I figure out what I’m doing wrong!”

I find that scheduling weekly meetings with my advisor helps. She has more experience than I do, so she’s able to suggest things to try that I hadn’t thought of.

My Ph.D work is solitary. But in other projects, I’ve found that working with a good team also helps me tell the difference. If I can turn to a colleague and say “Hey, I really can’t get this to work out right. Is there something I’m missing?”, then it’s much easier to keep taking steps forward and get unstuck faster. (And of course, they can turn to me and ask the same thing, so two people get unstuck.)

So I’d much rather work with a good team than do purely solitary work. I find it reduces stuckness.

Ph.D dance-off!

The results for the fourth annual “Dance Your Ph.D” contest, sponsored by Science, are in!

First prize went to “Microstructure-Property relationships in Ti2448 components produced by Selective Laser Melting: A Love Story”, created by Joel Miller, a biomedical engineer at the University of Western Australia in Perth. (Go BME!) He is working on ways to make better hip implants — ones that are more flexible, more customizable, and will last longer. (Click the link to his video on Vimeo to see his full writeup.)

Here’s his video:

Microstructure-Property relationships in Ti2448 components produced by Selective Laser Melting: A Love Story from Joel Miller on Vimeo.
 

The three other winners — smell-mediated fruit fly courtship, X-ray crystallography, and pigeon society — can be seen at the main Science article.

I guess this means I need to start choreographing my research for next year! Since it’s about arrhythmia, I might take inspiration from the classic Diagnosis Wenckebach, performed by the 2010 Med class at the University of Alberta.

Diagnosis Wenckebach

Goodnight, Mr. Jobs.

“All external expectations, all pride, all fear of embarrassment or failure – these things just fall away in the face of death, leaving only what is truly important. Remembering that you are going to die is the best way I know to avoid the trap of thinking you have something to lose. You are already naked. There is no reason not to follow your heart.”

– Steve Jobs, Stanford commencement address, 2005