A really cool way to make interactive web apps using R as the backend.
I can already see this being used in future lab meetings:
“Oh, you want me to try this parameter? Why, let me just change it in my web app!”
The potential use cases for this are fascinating.
The mailbox app has pretty much changed my working day.
I used to spend good amounts of time rifling through the wall of text that was my inbox, for no reason at all. Just going up and down the inbox, checking over subject lines, making sure I hadn’t marked as read an important email, and reminding myself of everything I had to do, important or otherwise.
My previous work flow with my email was one that required a lot of subconscious work. Anything that needed to be done, or remembered, or noted, or that was even mildly useful, was simply left as unread. Stars were occasionally added, but otherwise the “mark as unread” button was a close friend. The thing is, it took quite a bit of time everyday to look through that wall of unread emails and remember how important each thing was, and when it needed to be done, if it was even an action item.
Enter Mailbox.
Sure, the reservation system was a bit of a drag (and now will be even more so given its popularity for new users). But it’s beyond worth it.
Now, whenever I get an email I can do one of the following, immediately, from the comfort of my phone:
To say that making my inbox reach zero is a joy is an understatement. It is now my priority whenever I get an email, because I hate not seeing the beautiful daily “zero” inbox picture (it’s quite the subtle, but effective, reward for reaching zero).
And now, I don’t uselessly stare at my wall of unread emails for good chunks of time during my otherwise productive day, and am reclaiming my time from the dregs of email.
Zero is an under appreciated number, and the concept of it was discovered long after the number one, and it profoundly changed the field of mathematics. The case is the same for email – its taken a while, but zero has once again brought about a revolution.
Reclaim your day, and your inbox. Mailbox is my new app that’s here to stay.
Right now I’m working on the zip code data set from Hastie et al.’s Elements of Statistical Learning. The homework wants to compare various methods of classification, comparing k-NN, LDA, QDA, and logistic regression (RDA is optional).
To start, the data is organized as a set of 16×16 grayscale values, with the true digit value being reported per 16×16 set. The data is already split into training and test subsets, however I’m sure this could be mixed together so that we can test various groupings of training to test data.
Testing k-NN, since we have a tuning parameter k (the number of nearest neighbors to consider), I’m first doing a cross validation step against several different values of k to find the k which produces the minimum number of errors. Following this, using the best k, I then run the KNN on the test data, resulting in the overall knn test error.
The one problem I’m running into right now is that a supplied CV method for kNN (knn.cv) uses leave one out cross validation, which results in 7291 different runs of kNN in this case, per value of k. So basically, it’s taking forever and got stuck. Hence why I have time to write this mini-post on what I’m up to.
I digress. Back to coding!
And if you’re at all interested in really advanced statistics and data mining concepts (e.g., reading the language of statistics is like reading a 5th grade level book), then you should definitely check out the freely accessible book, Elements of Statistical Learning by Hastie, Tibshirani, and Friedman. It’s a great resource for aspiring data miners, and there’s tons of data sets to try out your data mining chops on.
It’s awesome that data, and with it, knowledge, is becoming more accessible than ever before.
Interesting piece on Seth’s Blog about working for free, and the value of doing work for free. The first lines really sum up the delineation between “work” and “hobby”:
“Work is what you do as a professional, when you make a promise that involves rigor and labor (physical and emotional) and risk. Work is showing up at the appointed time, whether or not you feel like it. Work is creating value on demand, and work (for the artist) means putting all of it (or most of it) on the line.
So it’s not work when you indulge your hobby and paint an oil landscape, but it’s work when you agree to paint someone’s house by next week. And it’s not work when you cook dinner for friends, but it’s work when you’re a sous chef on the line on Saturday night.”
Now, he goes on to talk about how work for free does not necessarily mean that it is work for no value. In fact, there may very well be work for free which provides much inherent value. The upstart band that wants more exposure, the artist that wants to promote themselves, the TV appearance, so on and so forth.
To go back to the very definition of work, it is something that you do as a professional – something that you can get paid to do, or that others would get paid to do. To commit to the idea of work, it is something that is created on demand. While it certainly can be something you enjoy (and in that case, your work is your career), it is nonetheless work by this definition. There are cases where work for free may not be worth it – consider the example Seth proposes, where a speaking role at an already overloaded conference – paying his own way there – would not create any significant value for himself.
In much stronger language, opportunities are bullshit. All too often, these opportunities to promote oneself cheapen the value of your otherwise valuable work, making you an interchangeable cog. And when handed the “opportunity” to do something, you have only two options – to give in or to say no. Another extremely valuable quote from another rant, “I Will Not Read Your Fucking Screenplay” by Josh Olson:
“I will not read your fucking script.
At this point, you should walk away, firm in your conviction that I’m a dick. But if you’re interested in growing as a human being and recognizing that it is, in fact, you who are the dick in this situation, please read on.
Yes. That’s right. I called you a dick. Because you created this situation. You put me in this spot where my only option is to acquiesce to your demands or be the bad guy. That, my friend, is the very definition of a dick move.”
Because as Josh Olson goes on to say, you have two piles of work – one that you do for good friends, and one that you do for work. To pick up an “opportunity” from outside of these, he says, “I’d be an awful person.”
Really, all of this can be summed up in this handy dandy chart that answers the question, “Should I work for free?”
Particularly with some recent experiences where I’ve overcommitted myself for little to no value is that I need to value my own work (or even hobbies that border on professional), and after reading these thoughts, the conclusion is this: learn how to say “no.”
Sometimes a polite no, other times a fuck no, but in the end, a no nonetheless. Because when you say yes to something, the cost of that yes is that you’re saying no to other things which may be just as, if not more important. And that just might make you an awful person.
When I got the email for ASBMB Hill Day, an event where scientists in ASBMB get to go to Washington, DC to learn more about the political process, I couldn’t apply fast enough. Now, at this very moment, it is more important than ever for scientists to get involved with our government, as critical funds continue to be slashed from our budgets which could be contributing in multiples towards our economic and public prosperity.
Below are my answers to the questions regarding the ASBMB Hill Day application, and hopefully they serve to highlight my position on the importance of scientists being involved with government.
—
Brief research description:
Given the importance of monitoring vaccine efficacy and responses to viral infections, it has become more critical than ever to understand the mechanisms by which acquired immunity develops in response to these challenges. To better understand this, I am currently working to develop integrative mathematical models – combining high throughput methods such as microarray, ChIP- and RNA-Seq, and flow cytometry – to characterize the features which define the acquired response. These computational models will be able to predict probable trajectories of the response from early time points based on characteristic signatures, and thus vaccine efficacy and viral infectivity can be not only better understood, but acted upon in the clinic.
Why are you interested in participating in the ASBMB Hill Day?
As science continues to progress at an ever-increasing pace, the breakthroughs that have come about in laboratories have the potential to deeply affect many people’s lives. However, these breakthroughs are often held back by steep regulatory processes, keeping many potential cures at the bench until the hurdle of profitability becomes obvious to biotech. This is a core concern of many practicing clinicians and scientists, as much potentially life-changing research becomes untranslatable. Thus, streamlining these processes to facilitate cross-talk between academia and industry, and lowering the unnecessarily high hurdle for clinical trials is a key concern of mine. Attending the ASBMB Hill Day will help me to better understand the process by which those on “The Hill” make laws, and more importantly, learn how scientists can become more involved with the lawmaking process to help promote the interests of not only the scientific community, but the public as a whole.
What experiences have you had in communicating your research / science in general to a non-academic audience?
Previously I have been involved with coordinating and executing several outreach programs based at the University of California, San Diego. There I was responsible for planning the outreach efforts of the Bioengineering Department towards the San Diego Science & Engineering Festival, as well as the national USA Science & Engineering Festival in Washington, DC, between the years of 2010-2012. Additionally, I was a member of the planning committees and an active participant of the UCSD Engineering community’s outreach efforts to local public schools, promoting science through carnival themed events, classroom outreach, tutoring, and building days on the University campus. These experiences have further cemented my belief in the need to better communicate topics in science with general audiences – making the field accessible and interesting to everybody is key to unlocking the successes at the bench.
What – if any – experiences have you had with your elected officials / the political process?
With regards to elected officials, I have participated in a few events held at UC San Diego with local representatives in learning more about the political process, as well as participating in local town halls leading up to the 2012 election season. Attending these events has made me all the more keen to learn about the political process, and as I have moved forward in my career, I realize that it is absolutely important for scientists to become involved with our government. No longer can we be silent audience members when critical funding continues to be cut – we must communicate the essential contributions which science makes to society, from the economic returns to the public good. To do this, we must reach the people who make these decisions, and maintain open channels of communication to ensure our contributions continue to be valued.
AMEZTECH 