You are viewing ledflyd

AlgoRhymthics   
03:50pm 09/12/2012
  I think I figured out why I like [eastern european] folk music/dance so much:



My brother shared this with me. 

If you think about it, this kind of music and dance is very modular and algorhithmic so implementing these dances as sorting algorithms make perfect sense. And you can get some idea of the complexity just by the length of the dance. Tempo of the dance is different? That's explained by variation in processor speed.

Insert sort uses a Romanian folk dance which is probably my favorite. Insert sort is also quite good.
 
     

(5 Hear my song | Sing along... )

 
internet famous   
08:57am 03/10/2012
  ... not even close. But getting re-tweeted/favorited by people I don't know makes me happy anyway.

https://twitter.com/ledflyd/status/253193138280226816

I think there's some potential for career growth, social recognition through the 'twitter-space' in science and biostats domains. Especially since a lot of science twitting seems to be sharing papers with very short commentary. Though, like blogging, it would take a lot of stamina to launch into 'fame'. I'd also have to stop posting about beer and personal shit too.
 
     

(4 Hear my song | Sing along... )

 
design flaw   
10:35pm 15/08/2012
 


A Brita filter built into the refrigerator door! Yes somebody actually thought this was a good idea and put it into the GE fridge in Ruchi's apartment the Atrium. There are several things wrong with this concept.

1. It's built into the door. The water dispenses via gravity and when it's full there's enough water pressure to flow freely. But as it empties, there is less pressure and since you can't tilt it like a normal pitcher, the speed of the flow slows to a crawl.

2. It's built into the door. This means when filling a glass you have the leave the fridge open allowing the cold air escape into a hot room, increasing electricity usage. This effect is compounded by point 1.

3. It's built into the door. The top hooks into special slots in the frame so you can't use that space for anything else (like another shelf).

4. It's built into the door. And due to the special hook system, getting this bulky dispenser without human-holdable handles is a pain in the ass, particular when heavy and full. The result is a lot of water spilt on the floor, the whole time the door is open (refer to point 2).

5. It's built into the door. An alternative fill system is to leave the brita dispenser in the door and fill with a separate cup. This takes multiple trips between the fridge and the door (refer to point 2).

6. It's built into the door and due to point 3, the space cost of having a regular water pitcher is even higher if fridge real estate is a factor.

Other than someone in a marketing department at GE, who would actually think this is a good idea?
 
     

(5 Hear my song | Sing along... )

 
sleep   
01:12pm 31/07/2012
  Seeking advice on calibrating my sleep schedule. Call it being in summer mode or something but I've been having trouble waking up significantly before 10AM (!) which, given my sluggish morning routine means I get down to work unacceptably late. Considering I now occasionally - routinely - commute from Ruchi's place on 135th, the problem is exacerbated.

And, for the record, the time I wake up seems invariant to the time I fall asleep. Occasionally, I fall asleep as early as 11PM or earlier and as late as 2AM (or later). And, unless I have an early morning commitment  (such as a lab meeting) I don't drag myself out of bed until at least 10AM. It's ridiculous because I obviously don't need 11-12 hours of sleep. I feel like I'm a sophomore in college again.

I don't want to invest in any sleep tracking tech (though not because I wouldn't be interested in the data). Basically, this is screwing over my productivity b/c I have that many fewer working hours.
 
     

(8 Hear my song | Sing along... )

 
sick & end of term   
06:55pm 22/05/2012
 
mood: sick
A week ago, on sunday night I stayed up until ~4:30 am cramming together my Machine learning final. The project & presentation was due the next day. I managed to force myself into a restless sleep, despite thoughts and ideas racing through my head, pumping me up. At around 5 am, though, I suddenly woke up. I jumped suddenly out of bed with my head mind racing, in full-on panic mode, while my head felt like it was swelling up into a balloon. My skin, especially my face, started burning up while hot knives were driven into my sinuses. My muscles started to ache and my arms spasmed. I slurped down some water and splashed some on my face and forced myself into bed, alternating panic with fever dreams for the rest of the night. Honestly, the only comparable experience was a bad pot experience, and at least that went away the next day.

In the morning, my head was hot and swam. fighting off vertigo and constant nausea and chugging water, I managed to type up my ML manuscript, took the train down to washington square (definitely no long walks), print it off at Bobst and arrive at class in the nick of time. Somehow I gave my 4 minute presentation without passing out. My heart was pounding for the whole thing and my head was swimming.

My best guess was (and is) a viral infection. I've been able to eat and while my appetite was diminished, no other digestive issues to speak of. So I figure stress left me vulnerable to some virus and that I'd get over it by morning. 

Well that was too optimistic, as it turns out. I spent the week in a near constant state of low-grade fever, muscle spasm and pressure headache. Around wednesday, I really started to panic. Could this be something really serious? What if i'm still really sick for my departmental works in progress meeting (which was today). There must something about constant fever or constant headache (or maybe just constant pain) that messes with identity and possible cognition (though simple tests suggested my mental functions were basically ok). I started getting depressed about being constantly sick, with no plausible relief and anxious that I wouldn't finish.

Finally dragged myself to student health the next day. The doctor didn't do much, but agreed with my virus assessment and reminded me to drink fluids with more electrolytes (not just water.. do'h) and take tylenol regularly (which helps a lot, but I don't like taking because it screws up my assessment of my symptoms).  I explained my extreme anxiety and asked about family history, but told me that constant dehydration due to fever results in adrenal fatigue and increased output of adrenal. This was a very useful explanation and explains why I could go to relaxed to full on panic mode on a drop of the dime. I was still anxious about being sick, but its nice to be able to blame physiology.

On friday, I dragged myself into brooklyn so Ruchi and her mom could take care of me. This, more than anything, really helped. Despite fever and headache and dizzyness and, in general, feeling non-human, it was nice having a support system. Firstly, knowing if something was seriously wrong, I wouldn't pass out and dehydrate to death, alone in my room (yes, i know it sounds hypochondriac but this was a serious concern). But hanging out with Ruchi was the best medicine I could've hoped for. She made me laugh when I was feeling anxious/depressed and feeling like I would never get better and made fun of my hypochondriac panics. They say laughter is the best medicine, and comedy has really done wonders for me, and in the past week has really brought me out of some depressive bouts and existentialist panics. But she was there for me and I couldn't imagine being alone and without her when sick.

On Saturday, my parents came over for a visit, which was nice. My mom brought her herbs and tinctures (which I wasn't feeling alert or cynical enough to debate about) but I did learn that I actually like mixed fruit/veggie juices and chia seeds in drinks. Meanwhile, I had to start working on my presentation for my big department meeting. I won't go into it, but I had a few pieces to add, so being out of commission for the whole time I was supposed to work on the project wasn't the end of the world. But the end result was not as polished and streamlines as I had planned. Also not as well rehearsed.. I finished the final touches about an hour before the meeting. This in addition to the adrenal fatigue had my heart going up to 130 bpm right before the meeting (my normal resting rate is ~60bpm). Presentation went ok, but I went over time and some people asked some dumb questions.

I'm not completely better. As I type this, the back of my head is throbbing and I still have some muscle twitches and dizziness, but I started to feel human again yesterday morning. Now that my meeting is finished, I'm looking forward to a stress-free night of sleep and hopefully I'll be back to normal tomorrow.

I do have a completely new-found respect [and fear] for people with chronic pain and infection.
 
     

(4 Hear my song | Sing along... )

 
looking for title   
01:29pm 30/04/2012
  I'm thinking about starting [back] up a science blog and I need help picking a title. The focus will be microbiology, particularly on current and perhaps historical research and especially DIY, at home stuff (particularly food/drink fermentation) with a flair for experimentation.

Some possibilities:


Read more...Collapse )
 
     

(9 Hear my song | Sing along... )

 
low cost exercise    
04:15pm 23/04/2012
 
mood: ho hum
Sitting for at least 11 hours per day confers a 40% high risk of dying within 3 years as compared to people who sit for less than 4 hrs, apparently even after adjusting for age, income, health status, etc.  Well, at least for Australians who participated in the study. 

 And while I don't completely get how you can control for such highly linked factors (such as idk, being a couch potato and excessive potato chip consumption) I do agree with the gist of this and have seen similar studies with the same conclusion. Like most of my generation I spend an inordinate amount of time sitting at a desk, pretty much on my computer (except for brief periods of class) from the time I wake up to sleep. While I actually do stay pretty active by walking to and from washington square park several times a week (and plenty of other destinations) this doesn't result in a sufficiently elevated heart rate. My attempts at working out have failed in the past because I don't trust myself to be consistent (present day Zach doesn't trust future Zach to keep up the effort, and effort today is erased by lack of effort next week).

But there are things that I can do:
my desk @ work

I don't need a standing desk at work, thanks to a conveniently placed shelf that I can keep my laptop on in addition to tall benchtops designed for bio lab research space. I'll need to come up with something else for when at home though.

Now I just need to not fall into the trap of substituting standing for a little more each day for proper exercise.  
 
     

(10 Hear my song | Sing along... )

 
tourism   
11:21am 16/03/2012
  I 'rescued' a bunch of tourists the other day. Coming home from San Fran back to JFK, I took the airtrain to connect to the E train at Jamaica. The signs on how to get around are really bad here. Its hopeless for people who don't speak english well, and I'm sure many tourists get stranded trying to get into Manhattan.

Anyway, I was getting into the subway with a bunch of confused looking [Dutch?] tourists, when I heard a shout... "hey if you have bags you have to go through the door." Not strictly true, but probably the Jamaica/Sutphin station has a lot of baggage going through it, so I didn't think anything of it at first. 

I walked towards the train and overheard a different [black male] voice "Hey man, leave them alone, you don't even work here!"

Curious, I turned back and saw that the original comment was not made by an MTA employee, but a non-uniformed and creepy-looking Indian guy. The tourists were halfway through the open emergency gate with a loopy-looking black guy hovering around with the older Indian man trying to appear authoritative (and failing).

"These are the wrong tickets, give them to me." The tourists, looking hopelessly lost and confused, handed the tickets over.

Uh oh. I walked over to the crowd. 

"Hey give those tickets back" I say, "you don't even work here."

"Neither do you, and they are the wrong tickets" he answered.

"Yeah but I'm not trying to steal their tickets, just give them back, man."

The guy shrugs and hands the metrocards back and walks away.

The tourists thank me, looking relieved... I know from experience, its a relief to have a friendly face when you're in a foreign place, someone who isn't trying to rip you off or hurt you. They are trying to get to Times Square and, after showing me the card, I realize the creepy guy was right... it's an MTA-LIRR combo ticket. 

"Actually that guy is right... that's a ticket for the commuter train which leaves from upstairs. It's faster to the city, but a longer walk to Times Square. As long as you're down here already, you might as well take the E train. I show them the Subway map on the wall, and meet up with another tourist whose staring helplessly at the map and asks me how to get to Broadway.

I explained how the avenues work, point out how Broadway stretches across Manhattan and point out where he needs to make the transfer to get to 101 street. During this conversation, the other tourists walk away, but I see them later on the subway platform... being serenaded by the rambling nonsense of the black guy that first grabbed by attention. 

Its a hilarious, but familiar, sight. These tourists first introduction to the city is getting lost in a dank subway, almost getting their tickets stolen and now getting bombarded by a well-meaning crazy person.

"Welcome to New York" I say as we finally board the subway.
 
     

(2 Hear my song | Sing along... )

 
I hate science, pt 42   
03:07pm 17/02/2012
 
mood: blah
When to an interesting talk today by this guy from Duke who's doing high throughput metabolomics to study metabolic diseases (particularly obesity and diabetes). It's really cool, cutting edge stuff. He's working towards integrating -omics data (genomics, transcriptomics, microbiomics, metabolomics, etc) to identify biomarkers for disease, predict outcomes and generate hypotheses. 

Aside for the presenter's awful and liberal use of comic sans in his slides, the talk impressed me and stressed me out. Obviously I can't expect, as a 2nd year grad student, to have as much to show for myself as a a tenured professor. But these days, learning about good science, especially in my field, is just a reminder of how far away I am from producing original science/computational methods. 

Science is so competitive these days, much of the low hanging fruit is gone, and there's the constant fear of getting scooped or just outmatched by the thousands of other people working and stepping on your toes. 

Of course, there's lots of collaboration and support (probably moreso than other industries) but it often feels like a backhanded, jaded, sink-or-swim collaboration (screw this up and your on your own). There's also the sensation of a Zeno's paradox of progress... as in, no matter what you accomplish in the short term, there's always an infinite amount of space left to traverse.

Which is why its easy to not do any work and watch tv all day... not that I do that.
 
     

(5 Hear my song | Sing along... )

 
i'm embarrassed by food   
11:02pm 08/02/2012
  I don't like to cook if my roommates are home. I also don't like them to see me bring in take out. It's a feeling I can best describe by "embarrassment". My best guess is that its a carry over from being made fun of for being a vegetarian and bringing in "weird" lunches in middle school. If I was a hipster, I'd want to be normal, but only ironically.  
     

(3 Hear my song | Sing along... )

 
lab meeting   
07:10pm 18/01/2012
  I gave a lab meeting today and, although I didn't know beforehand, a bunch of faculty from the bioinformatics department showed up! And to be clear, these guys are hardcore computer and statistical scientists, so when I saw them show up 5 minutes in, I immediately knew I prepared the wrong presentation.

Two basic observations, which I will keep short:

1) keeping presentations timely and informative is difficult when you don't "know your audience" beforehand. I had too much experimental background and work from other people that wound up bogging down the discussion. the Bioinformatics people kept asking technical questions that I couldn't answer (didn't know experimental and analysis details because it wasn't my analysis and I lacked technical expertise) so I would up rushing through all the important things and still ended 20 minutes late.

2) Keeping presentations informative is difficult when your audience has better grasp of the concepts than you do. 
2.5) the above is a problem because you can't really address a potential problem when you don't understand the question.
2.75) making up an answer is difficult but not impossible (and if you don't stumble the answer might even sound impressive to someone)

3)  There's a lot I don't know. Machine learning and  related topics in probabilistic graph come to find. Is it possible to cover the basics of stats and probability while "skipping ahead" to the hard stuff. I don't have the time or inclination to start taking undergrad courses in this stuff.

4) ultimately, people ask the tough questions when their interested and engaged in the topic. Sure they might signaling to others how much they know, but if they don't feel competitive they probably just don't care. 

I'm really doubting myself right now, but its also strangely motivating.
 
     

(5 Hear my song | Sing along... )

 
Resolution check-up:   
12:21pm 02/01/2012
  Lets see how well calibrated I am by checking up on last year's resolutions.

http://ledflyd.livejournal.com/46529.html


1. Keep better track of my finances (CI - 50% - 65% for 1 month, 40% for 3 months 30% for 6+months)

    I think I did this one for less than 1 month, and not at all after that.

2. Become a better rationalist (CI - 80-90% for the year)

    I guess I succeeded in this one, but I fail for being too vague.

2.5 Reach 2000 beer ratings by Dec 31st 2011 (CI - 75-80%)
  
     I was pretty confident about this one, but I think i've failed. I have about 1590 beers rated on ratebeer, but the big limiting factor is my big back log of beer notes. I have a couple hundred in my backlog, so I'm not tracking those by the numbers.

3. Read at least 2  extra-cirricular books per month  (CI - 60%)

   I haven't explicitly tracked this, but I'm pretty sure I am at or around this.

4. Become competent computation biologist by the end of the year (CI - 75%)

    Again, I suffer from being too vague. But considering at this time last year I didn't really know how to program and I had never taken a math/CS-based comp bio class, I've definitely gone a long way in tackling this one. Don't know if I consider myself "competent" yet.

5. By May, decide what the focus of my PhD will be (CI - 70-80%)

   If I didn't know by May, I pretty much do now. Although this is subject to change for at least the rest of the semester, due to natural forces.

6. Make at least 1,2,3 new friends (CI - 80%,70%,60%)

    I made some new friends but failed to make any close friends, and in fact I think I drifted away from people I thought I'd become close with. 

    I'm ok with how I did this past year, especially for the ones I knew I could do without working for (like 4 or 5 - Those were going to happen whether or not I was explicitly thinking about them as goals). I'll make a 2012 update ASAP.
 
     

(2 Hear my song | Sing along... )

 
I got scooped (kinda)   
01:13pm 20/12/2011
  Science is a cruel mistress. I read this paper right before going to bed (thanks to a blog post by Andrew Gelman), which of course resulted in me laying awake for hours worrying if my idea got scooped. Lots of things to talk about here, so I'll try to keep it organized.

Firstly, I didn't really get scooped, just the idea that I had to use mutual information to calculate associations between bacterial species in a large, micrbobiome data set. The method I was thinking about is completely different than the paper. So, it may be ok. I haven't done much work on this yet (other than a short script to calculate mutual information via simple binning procedure). This 100 line script actually took me weeks to write, but that was way back in last year while I was still learning R and didn't really know how to program. 

But that leads me to another point: I'm still a novice in programming and statistics, and the idea presented in the paper is actually quite sophisticated. I never would have been able to come up with anything like that on my own. Hopefully, this will provide motivation juice that will help propel me into the winter break, where I will attempt to do a lot of catching up. 

That being said, there are some good things: even the authors admit some weaknesses in the approach.. the problem with binning data is that you need some heuristic or regularization to prevent bins from becoming too fine (which would probably result in the maximum mutual information, but be meaningless). All this means is that the method is probably computationally intense. Additionally, this means is that the procedure is not scale-invariant, so getting the proper bin size can really effect interpretation of the results. They can detect non-linear relationships (through some modifications) but not in a way that will be intuitive to a non-mathematical practitioner.

Additionally, they don't address the impact of closure on finding associations in a dataset. In a compositional dataset, such as the microbiome, data is normalized to total counts, which results in the parts not being independent.  In a microbiome dataset, this is a [largely unaddressed] problem because false correlations or associations can be created by the dataset being dominated by a few abundant species, causing false correlations as those species vary (think about it this way: if there's a bloom in a species, what happens the the real population and relative populations in other species over that bloom).

The above paper doesn't address this phenomenon (which is brought up at length by Aitchison in mathematical geochemistry journals). I'm planning to test it by the "shuffle test": if you shuffle data around, by randomly putting actual bacterial counts into 'organisms' real associations should disappear. If an association arose due to effects of normalization, those associations will appear in a shuffled dataset, since the abundance information isn't changing.

People studying gene interactions on the systems-level are probably used to being "scooped" like this, since there are so many methods and procedures now. Microbiome is a new and "hot" field so there is less going on, but it's going to explode soon. In other words, I just have to get used to people who are smarter than me publishing more sophisticated methods. Meanwhile, there is still plenty of low-hanging fruit. I learned about compositional data problem from Eric Alm at MIT. They have a good method, though there are still caveats which could be addressed by people (like me!) without a ton of statistics knowledge.

And, in the end, I do work with a lot of smart people doing cool things in both the experimental and computational end. If I don't wind up developing a killer computational technique that transforms the field, that's cool. There is still a huge dearth of people applying proper analytic techniques to experimental datasets, and many statistical techniques have not been properly validated yet. And, I'm at heart a scientist, so if I come up with an interesting model, use somebody else's technique, and come up with interesting, new biology, that's still a win.

I also have some other ideas that I haven't seen (really at all) in the literature, so I'm still in good shape as far as my (eventual) thesis is concerned.

It's still just a little frustrating to know that you aren't as unique, innovative or smart as you think you are/ as other people in the field.
 
     

(5 Hear my song | Sing along... )

 
communicating math   
04:26pm 07/12/2011
 
mood: tired
I just got back from a class where I attempted to explain multidimensional scaling (which microbiome researchers call, somewhat old-fashionedly, principal coordinate analysis) and, after getting lots of blank stares, realized that I need to optimize for communicating mathematics. Since I'm going to be talking to biologists, primarily, and I need to sell them on why new statistical/computational procedures are needed, being able to talk about those techniques is essential.

Part of the problem is that I've never used MDS and only have a passing familiarity with the underlying math, (and I've also taken a necessary break from my online machine learning class). However, the above is also true for techniques that I AM using. Part of the reason I need to get a better handle on multivariate techniques, like PCA, MDS and DFA, is because of my belief that researchers are applying them inappropriately, perhaps ignoring potential pitfalls and caveats in their interpretation of the analyses (which is a minority, but not an original viewpoint). So I also need to talk about what people are doing wrong - which is arguably more difficult than just explaining a technique. 

Another pitfall is difficultly in conceptualizing mathematics. I'm very visual when it comes to understanding mathematical concepts, so if I can't form a concrete idea about it in my head, it will be impossible for me to explain to anyone else. This is a problem, because [for example] I have  no idea what the Variance of a logratio should look like. 

One of the ways I've found around this block is to implement them myself, either on paper or on code, which was Bobby's advice, but it isn't practical to recapitulate big projects. 

Would love some advice from the math people or in general on this. Ideas that don't involve going back to do an undergrad degree in AMS would be especially appreciated! 
 
     

(8 Hear my song | Sing along... )

 
goals, etc   
11:00pm 03/12/2011
  I've been promising ruchirahni I'd do this for a while now. I want to outline some of my goals, for organizational and feedback-requesting purposes.  But first a little meta-commentary:

A programmer friend/lab mate shared this article with me (Hidden habits of ineffective people) in the context of a conversation about a guy in our lab who [she thinks] is useless and has tricked our PI (Rich) into believing he is useful. The point though is that I think that I fail a few of these "usefullness" criteria, which I will outline and respond to. In the spirit of productivity and to be less self-pitying I will justify how the point doesn't apply to me or how I'm trying to overcome it.

1. Consuming more than you create - 
This one is tricky for me because I have a lot to learn before I start producing useful algorithms, data analyses or scientific content. On the other hand, I have a few good ideas which haven't been tried before (in my field) and there's no big rush to publish tomorrow. Some pure consumption is turning out to be a necessary first step (even if this first step is seeming like it's taking 'forever').
2. Watching your own vanity metrics - 
This is not a big problem for me. I know my limits and I have no problem seeking out people who know more than I do, flatter them and ask questions.
3. Starting the day responding to others -
This may or may not be a big problem. I spend a significant amount of time checking my email, but I'm not sure how much of this is to respond to other people's request and how much is just gathering information (albeit of varying degrees of importance). I do usually make a game plan in the mornings about what I want to accomplish, so it would be useful to keep track of to what extent other's impedes this. My guess is that I sometimes lose whole days due to other people, but that it isn't a pervasive problem. Also, I think my PIs should be somewhat exempt from this, b/c without processing their requests, I won't graduate! I think the original article inappropriately de-emphasizes the need to satisfy other people to follow through on your own goals.
4. Prioritizing the wrong activities - 
This is somewhat related to point 3, but the OP mentions the concept that busy work is useless and you shouldn't do it. As much as I'd like to get rid of busy work in my life, I still need to take classes (as much as it feels like busy work) there are some requirements I just need to fulfill, hoops I need to jump through, etc. I would love some tips on how to recognize busywork and how to avoid it (or at least make it useful).
5. Relying on multi-tasking to "save time" -
What is multitasking anyway? I have a couple of coding projects going on, some wet lab work coming up, programming homeworks and a ton of reading to do. Is it multitasking to rotate through projects on different days (or half-days), or is it more effective to work through something to completion before moving on? During a lab meeting, it seems like I'll have more to talk about if I have more projects going on, but this could just be less progress on more work.


Goals after the cutCollapse )
 
     

(10 Hear my song | Sing along... )

 
The hunt   
12:54am 04/10/2011
 
mood: pensive
I woke up at 6:45 this morning to go beer hunting. Actually, my alarm didn't go off by accident and I thought I would miss the Great Hunt, but thankfully the B train is running express again out of Brighton (as of today, after a two year hiatus). Ruchi's mom (who has my profuse thanks) dropped me off at Kings Highway Q/B station and I zipped across Brooklyn - only 20 minutes to Prospect park - and into Manhattan. I walked from the Broadway-Lafayette station to Bowery st. Whole Foods where a line was already forming for this release of Founders Canadian Breakfast Stout. I got there at 7:45 or so. 15 minutes before Whole Foods opens and managed to get to the number 8 position on line. Whole Foods only got 1 case of the beer, which consisted of a mere 12 750ml bottles. My buddy Matt (from NYU, who I introduced to ratebeer over a year ago now) rolled up right behind me, but at least a half dozen people were turned away.

 The bottle cost me $19. Not bad, considered the high demand. A friend of mine didn't get any, despite going to three different stores in NYC. After Whole Foods we went to New Beer Distributors on Chrystie St, but they claimed they didn't get any case in. The word on the street was that they freaked from the crowd coming in early from Whole Foods and decided not to sell any beer from their case.

I spent the next two hours at a lab meeting at my downtown lab (I'm organizing the meetings now) and I had to talk briefly about what I'm doing. Which is nerve-wracking because I'm woefully under-prepared to talk about the computational side of my project - but more on this another time. After, I spent the next half hour preparing an email to send out for my "2nd job" at SKI - a beer NYC beer distributor (the company happens to handle the Founders account related to the aforementioned beer).

Then off to Good Beer on 9th st btw 1st and Ave A, a beer store that opened at noon. I got there at 11:30 and was 14th on line. Damn - they only had 12 bottles too. So I go home and peruse the Ratebeer forums. About 20 minutes later I get a tip from a RBian about a new beer store/bar on east 29th st that got in 2 cases of CBS with a 2 bottle/person limit. I rush over there, texting Matt to tell him the good news. I roll up to get the last bottle in the store. Matt comes along the street as I'm leaving. Sorry, I gesture to him... the beer is gone. And I'm another $20 poorer (but 1 beer richer).

I go back home and attempt to work on my memoir for the week (the title might be "Taliban Beach") but I was too flushed with dopamine (?) to focus. I have a deal arranged to trade the CBS for some other beers, but the beer has bids in excess of $100 on ebay, so that's tempting too (though in the beer geek world, overpricing beer on ebay is a big Dick move.)

Didn't get much work done today (though I did finish a Java homework assignment while splitting beers with Matt, while he watched football) though I did get to deal with Claudia, an MD in my lab with a math undergrad major and a big heart and a big laugh, telling her how to Properly organize a spreadsheet.

I don't know why I spend time chasing after beer. This isn't the earliest I've woken up to chase down a beer and it isn't the most money I've spent on a beer. But it is probably the least 'rational' habit I have. The rush does feel good though. Better, probably, than the beer tastes.

I'm reminded heavily of when I used to volunteer at I-con, a sci-fi convention held at Stony Brook every year. I used to spend hours setting up the dealers room, setting up countless chairs and tables around campus, and then wandering around the con wondering why I was even there... Sometimes, the thrill of being part of something, hunting, cooperating and even competing with other people for a product, overshadows the actual (though maybe not the expected) rewards.
 
     

(7 Hear my song | Sing along... )

 
Memoir class assignment 1:    
02:55pm 25/09/2011
  Write a nonfiction creative prose piece about an embarrasing moment you thought you'd never live down - but did.

I was thinking about this for a title: 
Peanut Mathology

Think about the follow question: If Tommy has 5 walnuts, 10 hazelnuts and 15 peanuts and wants to share each type of nut equally between
himself and 2 friends, how many nuts will each of them have?

A conventional math teacher reading the answer from a conventional textbook in a conventional elementary school classroom will give an answer along these lines: 9 nuts each with the following breakdown – 1 walnut each with 2 leftover, 3 hazelnuts each with 1 leftover and 5 peanuts each without any remainder.

However, I was a kid with a biologist father, an agriculturist mother and stuck in a world of imprecise convention. So, when the teacher called on me to submit an answer I, of course, answered correctly: “4 nuts each... its a trick question because peanuts aren't actually nuts!”

I still don't know why I gave this answer. Didn't I realize that an elementary school math problem probably wasn't trying to make a subtle point about plant taxonomy to trick unsuspecting 4th graders into screwing up simple division? Perhaps I was trying to show off to the class; to demonstrate my knowledge about the botanical distinction between the true nuts and the so-called 'nuts' of culinary mythology (a group that includes, I have since learned, almonds, pecans, walnuts, cashews, pistachios and a host of others).

Now, I have never thought of myself as someone with a burning desire to be right: I don't need to crush teachers and students under the heavy weight of my superior intellect. I like being emotionally rewarded for a good answer as much as the next diligent nerd, but I'm certainly not a praise-aholic. That being said if, on that fateful fall day in the 4th grade, I thought my classmates would be cheering my insight and if I thought my teacher would praise my nuanced view of mathematics and classification then I would be sorely mistaken.

The teacher looked at me with all the wit and perceptiveness of a grazing cow and said “Of course they are... why else would they be called peanuts?”


In hindsight, the mind usually exaggerates humiliation. The laughter from my peers, which probably was actually only a few seconds of tittering, to me felt like great gales of shame sweeping over me. I wanted to hide under my desk, to turn into liquid and seep out of this chamber of embarrassment. I withdrew into a dark tunnel within myself shuddering and muttering, “peanuts are legumes, peanuts are legumes, peanuts...”.

From that day on, I was not known as the boy who didn't know what a peanut is. I was not taunted and excluded by my friends as a snobby know-it-all. I'm sure the event has long since leaked out of the memory of my teacher and classmates. But, to this day, I still cannot correct a teacher.


 
     

(3 Hear my song | Sing along... )

 
new class   
12:37am 18/09/2011
  Because I clearly don't have enough to do: Still it'll be worth it because I can't think about science all the time, and I expect it'll improve my livejournaling ability.


Writing Seminar: Memoir & The Moment


Facilitator:  Madeleine Beckman, Contributing Reviewer Bellevue Literary Review, Adjunct Professor - CUNY


Date(s) & Times: Mondays, 6:00pm – 7:30pm


Specific Dates: Sep 19, Sep 26, Oct 10, Oct 24, Nov 14, Nov 28, Dec 5


Locations: Smilow Conference Room A **(Nov 14 – Skirball 2nd Fl. Conference Rm.)


Enrollment Limit: 15


Description: The best memoirs connect with the reader at a basic human level. To accomplish this there must be some uncovering of whatever wisdom was gained by the writer, (however slight) from a backward glance. In this seminar we’ll look at how memory informs the present. Why is a particular moment important to you to write it down? How does memoir differ from autobiography and the personal essay? We’ll explore the uses of reflection and narration, the changing shape of memories, the ethical issue of involving others in your story, dialogue, description, and structure. Each class will include discussion, readings, and writing exercises.


 
     

(6 Hear my song | Sing along... )

 
Tuesday, 2AM   
12:28pm 16/09/2011
  So I'm not much of a computer or math person (traditionally) but when you're in an environment where everyone has even less experience than you, it sort of turns you into the "computation guy." Fine... but I feel the pressure, because I don't really [feel confident that I] know what I'm doing.

But what I do know is to when to feel bad for computer newbs. One of my bosses (ok, Claudia isn't my boss, but an MD at Bellevue, who's turned researcher. We get along great and we're working closely on a few projects). Almost every day she and a research tech have to collect fecal pellets and weigh 40-something mice. This generates a lot of data. 

I got the idea that I should do some database work for the project (it's a good idea to learn anyway). The problem is that we aren't using Google docs, or something sensible, to coordinate. NYULMC is a thoroughly M$ addicted place and so they use Sharepoint over their servers (and due to HIPAA issues, we pretty much have to use it to maintain a database. Ok, so I'm sure sharepoint has some sort of internal database software but I don't really want to learn how to use it. Since Sharepoint uses SQL in the backend, I figured I might be able to figure out a way to maintain a database over sharepoint using MySQL (which I'm somewhat familiar with). This is a technical challenge though (and I stayed up way too late the other night figuring this out) for several reasons: 1) I don't know much about networking 2) I only know UNIX commands, so I'm sure there's gonna be some compatability issues between trying to control MS software from a Linux machine 3) What little networking I know how to do is probably useless because the NYULMC network is stringently protected, but the IT department is very unhelpful.

Figuring out work-arounds was getting annoying, so luckily I was talked out of trying to do this by Francis (my friend at NYU who's a very knowledgable CS person). She's done some database management herself, and said to not even try: HIPAA + network + NYU MCIT = impossible task. Plus without the pre-existing skillset, it would be a huge time suck to learn and since it has nothing whatsoever to do with my thesis, then who the hell cares.

Still, seeing Claudia's excel sheets, I knew something had to be done. I taught her how to use functions and explained to her why you should stick to a 2d strict Column/Row format. Don't use colors to convey info on a spreadsheet. Don't plot each Date as a new column and pellet #/mouse weights as sub-columns. Makes the data much harder to graph and manipulate later on. Apparently, Excel 2007 has this handy "filter" feature which means that generating spreadsheets with subsets of data is very easy. I think this will be sufficient for their purposes, and just continue to use sharepoint as a remote server for file storage. Fine. I don't even think it'll be worth it to invoke MS Access.

As for myself, I figured out how to use Libreoffice Base as a front end to do data entry into a MySQL database. That's going to be very useful. Also I learned not to try to take on too much (nobody asked me to set up a database) and let people work with the tools they know before trying to implement something they might not be ready for anyway. 

Back to data entry for me.
 
     

(4 Hear my song | Sing along... )

 
   
11:43pm 06/09/2011
 
mood: tired
I think you can tell a lot about where I am in life based on my Google reader. Music, Ron Paul stuff, libertarian stuff, general and free market econ stuff is on it's way out and almost gone. Overcoming bias is backlogged. Don't check LessWrong anymore (though to be fair I was never a regular reader in the first place).

Starting grad school and reading Gene Expression got me back into science blogging. Also following some statistics blogs, some ecology stuff, microbes, fermentation and food. Never followed beer blogs (they're useless) but my activity on Ratebeer.com is going strong.

Back in senior year of high school, I joined scienceforms.net. I haven't been a regular contributor in years (I'm still a mod) but as of right now I've logged 8,746 posts. My blog (hosted by SFN) was getting a little attention when I stopped posting regularly. Who knows, maybe I'll get back into it.

But I doubt it.

Time is getting crunched.

Classes started today and, coming from ruchirahni's place all the way in Mill Basin, plus leaving late left me with little time to do work. Every time I talk to Dr. Blaser, I get another project to work on so I've been trying to avoid him. I've been helping out with mouse experiments for various people (esp. sacrifices), analyzing microarrays (gene expression profiles) from mouse livers (and trying to figure out something computationally interesting to do with them), and most recently, developing a math model of estrogen circulation. So I've been learning how to use bioconductor (R and lots of stats) with the future hope of being able to apply machine learning algorhithms to get useful info out of it. Compartmental ODE models and linear algebra (to solve complex systems of equations) while researching model fitting and deconvolution methods. I just don't have the background for this! Additionally, I want to re-visit Mutual information as a method of infering causality of mouse adiposity phenotypes from microbiome data. That started as a project last year, and it has real potential.

Meanwhile, at least my PAC course is starting off with the basics. The professor is very good at explaning concepts: though my major complaint is that he's letting students push him into more advanced topics by letting them asking useless (for now) questions (the class is in Java).  I'm not sure if my fellow students are just asking to signal how much they already [think they] know, but I wish they'd shut up and let me learn. We're gonna have a hard time convering two semesters of undergrad material if this keeps up.

With the PAC course having 12 hours of homework a week, I'm not sure how I'm gonna manage once my Med Micro course starts up in October, or Works in Progress seminars in a few weeks. 

Being in two labs means that I'm always going to be over-commited, but I'm sure I'll figure out how to balance it all.

Anyway, the point I was trying to bring up in the beginning is that every few years I seem to have an "identity crisis" and my interests shift. Or, rather, I tend to get passionate about something and it swallows up my life. Ron Paul blogs 'consumed' duing 2008 (directly leading to some shitty grades). These days, you're more likely to find me on an extended lunch hour at Blind Tiger, getting rare ticks from some obscure brewery.

Of course, I learned my lesson, so I don't plan on letting beer (or anything else) get in the way of school. I'm very lucky, in a way, that the focus of my school/work happens to be something I'm so passionate about, and that I can practice progamming in the lab, and technically not have to feel guilty about it. But still, some interest has to make room for another b/c there's only so much time in the day.

When I graduate, maybe I can find a job making a lot of money doing computational microbiology to study free market fermentation and have a side gig playing eastern european folk music.
 
     

(12 Hear my song | Sing along... )