zero crossings

hi.

so it's about midnight and i'm in my office on campus. it's hot as hell in here, as the building has been waiting on an AC part for the last nine weeks or so. we have fans everywhere, and no papers stay still in this room. always fluttering, paper weights are everywhere and everything. it has been a busy night, debugging some old code bases, homework that was assigned friday that is due wednesday, install a new os, edit a paper that we've literally been editing for the last three years, and gaze out the window. check out my screenshot from tonight.

everyone tonight is obsessed with the big unc game. i suppose right about now chapel hill is blowing up as they usually do when they win the tournament. i never really got into it, but i suppose if i were a unc student i'd feel differently. at least we'll see the sports coverage level off here in a few days.

the to-do list just gets longer with every day. i now actually have three to-do lists: critical, soon, and stuff to do when i think there is nothing left to do. that's a lot of stuff. i'm guessing some things will never be crossed off. it's like a triage for my life. speaking of triage i finally got around to watching the series finale of ER last night. i'd been watching almost since the very beginning and i am True Believer of that show. i cried and i wept through the nostalgia that was the last episode. i'll miss that show.

i can't believe it's only tuesday tomorrow. so much left to do. i have a test next wednesday, and a presentation to give in the class whose prof i said the stupid thing to last week. this means i have to really, really vet my presentation and get this stuff pin point accurate or hell will break loose when I do the presentation.

this past weekend was the full frame documentary film festival. films you need to watch and that will change your life that will be coming soon to a theater near you (though most likely not): Food Inc, Sweet Crude, Shouting Fire: Stories from the Edge of Free Speech, Salt (absolutely, incredibly beautiful). once again, the festival answers the question "how much reality can you handle?" the answer, once again, was "a lot less than you think."

today i asked one of the stupidest questions to the wrong person.

i was going over the proof that newton's method converges quadratically if you start close to the solution, the derivative is nonsingular at the answer, and that the derivative is nonsingular. the question i ended up asking was essentially about the fundamental theorem of calculus (though, in a weird form). people, it isn't called fundamental for nothing. i was basically asking him to explain calc 1 material to me, and he wasn't having it. honestly, he chewed me out appropriately. it was the wrong question to ask to the man. man oh man, he put me in my place quick. it was so so embarrassing, like i'm surprised he didn't go tell campbell to kick me out of the program. i knew i should have asked dave my questions (he is also in the class, kinda pays more attention to things than i do).

anyways, i spent the next hour going over every detail and convincing myself of everything. there is just one final detail...this one little inequality that has to be worked out.

it seems like today has been about trying to understand stuff that i should have figured out a while ago. this morning i was trying to re-run some old optimization codes back from when john the former grad student was around. man, it was rough. it really seems like i'm always playing catch up, but i bet a lot of people feel that way in their job.

i got my HD cable box today for my new tv, which is great.

i still haven't found a job, and in fact i just txt'd with danny, the bartender at frazier's where i used to work. i think i may have mentioned how lamentable it was to be fired from the place, especially when the circumstances were arguable and all i did was play the hand i was dealt the best that i could.

oh well, the beat will go on.

winson pointed out that whenever i say that i'm back to lj it lasts for about five hot minutes and poof i'm gone. i suppose that's true. it's just been so hard to find time to do anything lately, even three minutes to pause and say what is going on in my life, if for no other reason than to be cathartic about it.

perhaps i shall take those moments now. a few weeks ago i was let go from the restaurant where i worked, which was really quite lamentable. i enjoyed working there and i'll miss the people that i worked with. it's been an odd thing having friday and saturday nights free again and having to worry about what to do. you could liken it to going to a school for years that makes you wear a uniform and then suddenly switching to public school and having no idea how to pick out what to wear in the morning. something like that. i put in some applications to some places, but you know these days...no one is really enthusiastically hiring.

school has been kinda weird, it seems like research has slowed down as i wait for my advisor to pick up. there isn't even really a light at the end of the tunnel, at least wrt to the semester. as soon as the semester is over, summer is here, which means intense qualifier study time. so in that sense, summer will be even more intense than the semester. studying my ass off for the three hardest tests i'll have to take. moreover, i can't believe it's almost summer. i can't believe that a year ago i was up in boston living it up and working at mit. that will be my *next* summer again.

no blips on the radar on the dating scene. haven't met anyone that i like, and i certainly haven't met anyone that really sparks anything inside of me. maybe it that i'm difficult to impress or that i have too high of standards, but i think there is just a gigantic dearth of {cute, smart, downtoearth, havetheirshittogether} guys around here. which i think is actually weird given the number of schools around here. yet i can count the number of gay grad students i've met on one and a half hands.

i've also been thinking about building a new rocket. a big one. maybe 5 or 6" diameter? it has been a while since i've done anything like that, like probably since i was a teenager. i found my old electronic packages and rocket box and i was thinking of how cool it would be to start a project like that again. of course, it comes down to timing, but i'm told that somehow we always find time for the things we love to do.

well, that should about do me for now. in all, i'm doing ok. if anything, i'm anxious and nervous about the summer and how i'll do on it. i can't let that anxiety work against me...i need to use it constructively. as for guys, well....i don't even know what to say about them anymore. one thing i am really looking forward to is the full frame documentary film festival next week in durham. every year it gets larger and larger. should be great. i can finally indulge my artistic and creative side.

until next time..

i leave for san fran tomorroww!!!! cold, wet, AWESOME san francisco!

i'm sure i'll be tweeting the hell out of it while i'm there. (virga on twitter)

maybe i'll get to see some of you.

love,
adam.

yay! i bought my ticket to san francisco today. i'll be there from the 1st through the 8th. i'm looking for things to do that i haven't done before, seeing as i've been to the city a couple times before. i mean, i've walked market street, i've wandered through {haight, castro, mission}, all that stuff. i need new stuff to do! let me know if you have any suggestions. i'm not that mobile, as i doubt my cousin will let me borrow his jeep and i hear the zipcar isn't really worth it.

it's been a long day. i've been doing...math all day. i take my homework really personally, so when i can't get a problem it basically haunts me. like, i put all my books away for the night and i'm still thinking about that problem...and i can't do anything more with it until i go talk to the prof for advice tomorrow morning. ugh. ugh. this class is frustrating in another aspect because the other students are really adolescent first years. they don't start the work until the last minute, they copy all of each other's work, and they just don't take it as seriously as i feel that they should. needless to say, i don't work with them. and this is a fucking phd-level math course in differential equations. you have to be so tall to ride the ride, and they are definitely too short for the ride they are on.

i should note that just sitting here, doing nothing, i can feel the muscles in my neck twist and turn. ow.

man oh man, so i just trimmed out my friends page and stopped following all of the lj_* communities that i was so into years and years ago. now my friends page looks manageable...i see actual people posting.

last night was valentines day night at the restaurant, and it was batshit-balls-in-your-face insane. i made the most money that i've ever made waiting tables, and it was a lot better than it would have been if i were at sullivans. i'm told i should save my money, which while true is also moot: i owe a couple grand in taxes this year (thank you waiting tables), plus i'm going to san francisco in a few weeks (note: buy airline ticket)..so the money has just gone out into the ether.

nothing ever changes. i'm still in school. i'm in the second half of my second year in a five year program. i'm still waiting tables. i live in the same place; i still don't have time to take pictures. in fact, i have rolls of film here that are waiting to be developed. oye. in the relationship department, still the same. still emotionally barricaded behind some false hope of something working out with someone from the past, thereby not allowing anything new to grow and develop. is AWESOME.

man lj, it's been a while. i guess i should update my profile to make it seem like i'm not a freshman in college anymore. at least change my default icon. i like my layout, but i'm sure if anyone sees this its because they forgot i was on the friends page. ohhhh lj, do people even use you anymore (who aren't in russia)? look at me, longing for the heyday of livejournal. oh well, it's nice to be back.

i have to explain why i am so frustrated working on this filter code.

i have two filters, the unscented and the extended filter. both come in continuous and discrete versions, and i'm trying to compare the performance of the filters on different models -- eg does the discrete filter paired with a pre-discretized system work better than a continuous filter paired with a continuous model and then sampled after the fact. obviously, this requires four separate filter codes, each slightly different from one another.

well, these filters have personalities and their failure modes are very hard to understand. sometimes, you may just need to use more data, or a smaller step size. or you may just have to tune the three covariance matrices, or the initial iterate. these are easy tweaks, but the failures that result from these look precisely like larger, more significant, logic errors in the code. perhaps most frustrating is the notation in the papers that is inconsistent, poorly explained, or just damn wrong.

i've been at this coffee shop for the last six hours coding this stuff up. i got so lost in the code that i just ax'd everything and started over, in a much cleaner and smoother way. i'm using the inv() command to take the explicit inverse of a matrix instead of solving the associated linear system. it's bad numerics, buts it so damn easy and all of my matrices are well conditioned that i'm aware of. dr. chu would kill me if he knew, but once i get this code stable i can add features.

i am on the last, most difficult code now -- the unscented kalman bucy filter. all of our models are continuous time models (duh) so its important to understand how pre and post discretization affects the convergence results.

and that is why i'm so frustrated.

i'm thinking that i should turn this blog thing back on. i tried to flip the switch a while ago, but never did. lately, i've been having thoughts that i wanted to put online but twitter was too short and the super secret blog out there was well....too secret.

so we shall see.

i also had the idea of starting an [applied] math blog because there isn't really one out there already, but that'd be a lot of work at a time that i don't really have a lot of time :(

standby.

made it into boston. here for the summer, staying with sam for the first few days. today we're beginning our brunch adventures and then going to ikea to get some crap that i will invariably need for my dorm room. it's kind of dangerous to do that before i've seen the dorm room but that's okay.

my friend adam e is up here to go to school at harvard, so we're hanging out this week. i'll update with pictures, as this requires visuals.

besos,

so i think i have some stuff to write about, finally.

these last few days have been kinda nuts, since having gotten back from boston. i had dinner with jim neal on saturday night which was a really delightful experience. he gave me even more insight into how fucked up politics is here in this state and about things we can do over the next two years to really build a strong network of people that can really transform the status quo in north carolina. i'm all for that. he's a very smart man, and it was good to have some 1:1 time with him to hear what he thinks.

i have but three weeks left until i head up north to intern at lincoln labs, and i'm starting to get nervous. i'll be working on discrimination and radar tracking of objects in the sky through real time 'parameter' estimation -- i think. i'm worried that i'll literally won't be able to do the math, and that i'll finally be found out as the math fraud that i probably am. (A's in grad school! hah! don't mean nothin). in these three weeks, dr. tran has me doing a lot of stuff. we have a cardiovascular model that needs some analysis (sensitivity, subset, kalman filtering) which be a nice set up to the computational methods that i'll be using this summer. the REU is starting up this summer and dr tran wants me to get the electron design group going and show him how my code from the last two years works. sweet. and i need to go shopping. business casual? the horror. let me wear jeans and a polo and we'll all get along just fine like. i'll wear khakis for the first day to scope it out and move on from there.

let me tell you what i'm really excited about: steak. i was at the farmers market with john today and i found a pair of farmers from near rocky mount that is selling all natural, grass fed, beef. after reading the omnivores dilemma this is awesome. it's local, sustainable, healthy beef. i can't wait to grill these guys up and taste the sustainability. mmmm mmm good. they were expensive, but they're actually cheaper than the grain fed stuff in the stores when you remove all of the subsidies out from the sticker price. speaking of money, i spent a lot of it this weekend. and when you add boson into the mix, we're talking some real bad money sinks lately. so much so that dare i say i miss sullivans. i ran into an old manager from there who happens to be one of my neighbors, who works at a catering business and he told me that if i want to work some jobs over the next three weeks to let him know...and i just may take him up on that. cash monies, baby.

and lastly on this noncontiguous update....lets talk tv. sad endings: the numb3rs season finale. what will charlie do next?! the ER season finale: man i hope it was sam in that ambulance that blew up. current hells kitchen: those people are fucking crazy. no kitchen is like that. lost: no matter how much you abuse me, i will always love you.

other things i wanted to talk about in this post but eventually ran out of steam: some of my HIV related research and things i learned at the optimization meeting; some more stuff on farming and young people that i saw in this article (stupid white hipsters!)

from my end of the wire to yours ---

so knowledge says that ER is coming to a close. tragedy. i will weep a thousand tears, esp because: carter is coming back. lets go africa. and kovach is leaving. double sad. weeping sad.

so i think i may go to cambridge and scope it out. there isn't anything going on here worth sticking around for between 11 and 3, so that may be enough time to go check it out. i'll consider this a dry run for when i'm here for the summer.

time for more math....this morning, its all about derivative free optimization. whohoo

So i'm up in boston this week for the SIAM meeting on Optimization. I'm presenting some of my research that i've been doing over the last few semesters....should be fun! I'm here for the week, then back to raleigh for a few weeks, and then back up to here for the summer working at MIT Lincoln Labs in Lexington, staying in Cambridge. Turns out my friend Emily is up there for the next few days, so I found a gay bar for us to go to tonight and we're going to go out and pick up guys. tee hee. hopefully the boston homoqueer crowd will be nice and inviting... There is a team of grad students here from Sweden that are just hot as hell..hot and exotic and oh so european. Isn't everyone from europe heteroflexible? I should exploit that.

So I successfully completed my first year of grad school. My lowest grade was a B+, my average grade was an A. Huzah! Here's looking forward to next summer. 

Jim Neal lost this past Tuesday. I have so much emotion wrapped into that race. He lost at the precinct that I was working at by 23 votes. Talk about being hard to move on from.

Sorry I don't update more even when I tell the world I am. 

...ok! time for the poster presentations!

anyone in boston until tuesday that wants to do anything? i'm here by myself and i'll just be walking the streets....let me know!

good evening world.

this week is exam week, and then i can hopefully say that i've successfully completed my first year of math grad school. that's got to be worth something i figure? on thursday i have my control exam, which i'm studying for all week. i think i more or less finished my numerical analysis final exam that was a take home. there was one question that was a thorn in my side all week long. it turns out it was from an old qualifying exam, and i have no idea how i would have ever been able to figure out how to do it during an actual exam. but i get to worry about all of that next summer.

i mean, there really isn't anything up. i can't wait to get everything over with. michael is subletting my apartment this summer while i'm at MIT, so that's a good thing. i'll be in boston for a few days next week for the SIAM meeting on optimization, so i'll give boston a dry run then. I've been neglecting my research like whoa these last few weeks, but I figure that if they really wanted me to do research they would give me research hours.

i am so awake right now. i took a 3 hour nap tonight and then had three shots of espresso in mocha form. hello nighttime anxiety!

...jay clifford is coming on friday and saturday!

i am so random. lets power through the week.

Let it be known...

Today is the official day of remembrance and recognition of the Armenian genocide. Here are some nice tidbits:

Armenia vows to increase recognition of the Genocide Keeping Turkey out of the EU would be nice.
Anti-Turkey protests on the ground in Yerevan. I would be on their side, obviously. Someone give me a torch.
Knowledge says there is going to be a concert in Yerevan soon.

Bush gives a (legally required) annual speech about the Armenians at some point today. He sucks more and more at it every year.

Keep fighting the good fight...

right so, i just bounced out of the award ceremonies for the math department -- i mean, they're geared towards undergrads and sure as hell wasn't winning anything from the grad students. the thing with the merit based awards is that they really don't....filter through to the people who could really deserve them. from the top down, they stop at the top. no one who isn't at the peak of the grades, or anything like that.

this kind of applies to grad school admissions. for instance, i'm pretty clear that i would be doing good work at harvard or caltech or something like that. grad work is grad work, esp in mathematics. i mean, linear algebra is linear algebra no matter where you go (with the notable exception of being taught by the late gene golub at stanford --- that man was linear algebra). but of course, i graduated with a 3.2 gpa from ncsu so those big schools wouldn't touch me. nevermind that for the first two years of undergrad my 5 year plan consisted of me bouncing out to san fran and becoming an emo-hipster-documentary-photographer. so the system doesn't really filter down to people who really could be great researchers, but was just having a "omgwhatamidoingwithmylife" crisis the first two years in school. now, i'm not bitter about this (though caltech...mmmm)...i'm just a realist. i'm going to apply for the DOE computational grad student fellowship next year which is right up my alley, but it's very competitive with people who are by any metric, better than me.

my point is is that i was never a math honors student, because i was never invited. i doubt i would have done it anyways...weird kids in general. but i'm still smart (well..eh) and talented (somewhat). i'm not bitter; just slightly jealous. all i have to worry about now a days is just the standard "am i good enough?", or worry about not understanding seemingly very simple mathematical stuff (backwards differentiation methods/implict eulers method, for instance). and there are always qualifiers to worry about. always.

anywho, it is the back half of dead week now. all that is left:

-- numerical final exam. at best, i am 50% done with it. at worst, a 30%. the hard question still looms (singular differential algebraic equations, if you're interested)
-- control final exam. next thursday. pwnage will occur. prepping for it this weekend on.
-- final paper in ralph's class. this will be straight forward i hope, but incredibly annoying.

just
keep
pushing

so this is a first for me. a lot of times in math and modeling, simulations are run thousands and thousands of times at slightly different values to see how the model reacts over large parameter sets. it's called a monte carlo simulation, and it allows a better sense of how the model works.

for my last project in my model validation class, i'm running a model that simulates a vibrating beam 10,000 times at very slight changes in parameters, so i can estimate the error associated with the model as compared to data that we have. it's kind of a neat thing, if i understood everything that was going on. in the past, we've never worried about the random-perturbations in the system...everything from someone bumping the table, the 60Hz power frequency, and cosmic events. we presumed that everything happened exactly the same way every time, no matter what. a convenient mathematical fiction as they say. you can obviously see the problem with this, since there is noise and randomness in nature and nothing happens exactly the same way twice. unfortunately, working these random-events into your mathematical model is difficult (though very doable) and requires some good background in statistics, which i can't claim to have.

as I was working on this project over the weekend and miserably failing at the stat part, i came up with a big list of questions that i was going to ask the prof today.....except he is in boston running in some famous "boston marathon". but i got a better idea, and now my computer is running 10000 model evaluations of this thing. so each model evaluation takes about 5 seconds to run, and i'm doing 10000 of them...i think i'll see the results tomorrow morning when i come in. i have an 3.0 ghz 8 core mac pro, and each core is working at 100%....according to top, matlab is taking 766% of available processor time. yay for computational power. i can only imagine how long this would take on my computer at home.

oh, i had to call GE today. we have a proximity sensor in the lab that can sense super close vibrations to the sensor, on the order of micrometer displacements. we wanted to get an accuracy estimate on the sensor but couldn't find the data sheet online or in the lab, and since it was no longer made i just made the leap and called the number printed on the device...which then put me to the main GE (they had bought the company) switchboard. incredibly, after about 5 different transfers and two emails, i got the data sheets for this damn thing. turns out, the error is about an order of magnitude greater than our data...which is....unexpected. more on that later.

i thought this morning it would be a good idea to walk to school since it was nice and cool and not raining. energizing to do this in the morning, utterly draining to do this in the afternoon. why can't people offer people rides on hillsborough st down the long stretch to morgan st? i mean, we're all going the same direction. tonight will be another night filled with working on all the shit that is due in the next week or so.

in the few minutes i have before numerical i'll note that its dead week, the week before exams. they call it dead week because well....you die. i've already forwent a trip to CA this week for IVEC '08 because of the workload that I have going for me here, including two projects, an incredibly hard numerical take home final, and more material in control. i'll only have one in class exam, and it's gonna be h4rd.

this past weekend was quite nice, just studying. oh, i forgot to tell the internet that five weeks ago i quit the restaurant, where i had been employed by the Man for the last three years. so long, suckers.

i was talking to johnathan a while ago about starting a math blog. i tried to get some other grad students on board in the department, but they were like "i barely like the math i do, how could i write about it in a way that others would like it?" fair question, i suppose. so it looks like it's on me. maybe once a week i'll write about some topic that i've come across that is too cool to not keep to myself, and i'll break it down in a way that momma could understand. and maybe i'll even use tags. i feel dirty.

ok folks. i think i am going to start blogging/journaling again. i put a new theme on the journal to make it look more blog-like, and we'll see how it goes. i need a new title, i think it is time to retire echoing vistas. i used to pull titles from lyrics of jump little children songs so i'll listen to their discography and see if anything stands out to me. i was thinking maybe something mathy, like "the next iteration", maybe i need something more dorky.

hopefully i'll be motivated to write about the standard faire that a grad student thinks about...math, hopes and fears, anxiety of failure, how crazy my research advisor is, and of course....debauchery and boys.

its good to be back, lj universe. time to put my permanent account in gear.

p.s, much to Jonathan's inevitable dismay, i started a twitter. the end times are near. but i still hate web2.0. i just thought it may be good to have so that i can txt it to let the world know just in case i am arrested in a foreign hostile country.

why should i care what robert scoble thinks?


i hate the idea of "tech culture". and blogger who blog about technology. i think the other bloggers are ok. but those stupid bay-area-silicon-valley "tech culture" blogger weenies are really dumb.


edit: and twitter can die. as if i care what you're doing at this moment.

man, i'm bitter at the internets today.