shoutout to MIT, my family, and my friends

this post is a shout out to MIT and the wonderful people here

i've had to incomplete a lot of classes this semester for personal reasons
but I'm actually really grateful because what it's taught me is that people are wonderful and trustworthy.

my hallmates and friends came to visit me, even though everyone is super busy with classwork and i didn't want to admit to myself i was actually in a life crisis
i gained new friendships with some incredibly strong people
i gained deeper friendships with friends who had gone through a lot, who could empathize with my struggles

my parents turned out to be adults i can trust and confide in, people who came to my side immediately when i needed it

the GRTs and housemasters, my medlinks, even my UROP supervisor, as well as MIT Medical and the S^3 deans have all been actively supportive

my professors, especially for small or new classes that are more flexible, have been amazingly accommodating and sympathetic
i've learned that the professors and staff think of us as people, not as the mythical genius MIT student who thrives on intellectual knowledge and has zero life issues
actually they've completely blown past my expectations in this regard. i don't think the struggles i face are that much more severe than the struggles other students here face, and everyone manages fine.

but i've come to be
okay with who i am

even more so now that i know that my friends believe in me, my professors and staff believe in me, and my family believes in me.

overall, i'm a really lucky person, and the world and MIT is a surprisingly wonderful place that confounds and delights me everyday.

===
other things i learned this semester: it's better to consistently show up for class even if you don't consistently get your work done before class, communicate with class staff, don't be afraid to ask for help whether it be for life or academic issues, don't be afraid to commandeer time in office hours even if you don't need the class for graduating (actually, you're paying tuition for it), resources are less strained than i think, it's better to go to class even 30 minutes late to show that you're trying than to give up and go to sleep.

more later, perhaps, as i think of things to add

Class Homepages Spring 2012

This page links to my class's sites to save me some typing (will serve as my homepage) [posted 12 Feb 2012, 8:48 PM]

DISCLAIMER to all my friends worried about my health: I don't actually do half of this stuff. I'm a cheerfully laid back sleep-when-I-want caffeinate-when-i-need completely mentally sane / average person.

Agenda / Calendar
PNG Schedule
old schedule

Great success! deleting blogger posts from google reader cache

So I accidentally this blog post here instead of on my private blog, and I immediately remembered being perplexed by how to delete test posts from google reader's cache a few years ago. It turns out you can't delete posts, but you can overwrite their content. Except I already "deleted" the post from the blogger.com frontend, so how the heck do I edit the content of the post? @__@

Now that superuser exists, it's superuser.com to the rescue!

I found this tutorial:
http://crsouza.blogspot.com/2010/03/partial-least-squares-analysis-and.html

Sort of linked from here:
http://superuser.com/questions/56908/remove-deleted-posts-from-central-google-reader-cache
Except the superuser link is to a ruby script, and it was much more useful to just do it by hand when I just had my most recent post to "unblog" instead of bothering with trying to install ruby / run it on windows *shudders*

I just changed the "post options" timestamp to 10 minutes later,  refreshed google reader, and voila my private content is gone and instead you see... this post :)

evaluations, sexism, startup cofounders

what are?

In 21w.789 I was talking to a classmate about her project and asked her, "how was the localization implemented?" She said, "I like interaction design, so once the backend is done, I'd like to work on the GUI. I know, playing into gender stereotypes..."

Hi guys. This is the universe we live in, that we cannot simply feel secure in who we are and what we like, in who we are as individuals, but instead feel compelled to be Women who are going against stereotype or following the stereotype -- but either way, the stereotype exists and defines our choices.

I learned the interesting fact yesterday that every semester in one of my course 2 classes, accusations of sexism or hitting on people come up in anonymous evaluations. (I was surprised -- I've never thought of TAs or UAs being sexist, I've always seen them as human beings and possible role models, perhaps fulfilling or not fulfilling gender stereotypes). I don't have any data on any other classes, but it's fascinating that this issue is on top of other people's minds as well, false accusations or not. It's that process sitting there chugging along eating along 5% of my brain cycles all the time, self-conscious, analyzing. (Yea, I'm pretty narcissistic like that, I never really get around to analyzing if people I know are sexist, but I sure do get around to wondering if I can blame all my failings on gender stereotyping :P)

how to not?

I feel that this is a societal level thing. Yes, it's good for you guys to open your eyes and ears and listen more and try to understand. It's good for you to stop and think: when you take charge of the coding portion of a project and do lots of work, you might be putting your friends in a tough spot where they may greatly appreciate you getting the work done but feel put off by your not giving them a chance to be "counter-stereotype". But there is no need to obsess about how not to be sexist on an individual level, I feel, since life is a grand experiment and trying and failing is acceptable.

startups

Yea, I think about gender stereotypes when looking for a cofounder and reacting to people's suggestions as well. I have a somewhat irrational determination that I will not work as a cofounder with a guy (possibly even a girl) who won't shoulder at least half of the "nontechnical" work in a startup. I would rather go it alone than risk the possibility of confirming gender stereotypes (yes, again, my choices are dictated by these stereotypes floating around in society. I don't think I am confident enough to try being an individual instead of a Woman).

from an email i sent recently:

okay, so my knee-jerk reaction to this statement [...] i will explain that it's because i'm not confident enough about my own technical skills to make me willing to shoulder the businessy work, as inevitably it will then take priority over technical work. in addition, i am paranoid about any possibility of contributing to the stereotype that girls are more suitable for nontechnical work. exceptions being made if the target is what I consider underserved markets, then i care enough about the end goal to put my own fears and doubts aside to contributing to making something happen
since i have a decent internship i feel slightly more confident about my technical skills, i am willing to help out with the business-y stuff if you want, but prolly not enough to make me willing to be a cofounder or work long-term this way -- also why i will only ever cofound a startup with someone who is willing to at least equally share the business work, possible small concession on >= 50% share of business work being made if cofounder is female.

tools for scientific posters, braindump

[Edit 16 May 2012: Please see my other post for a writeup and example, with svg source file, about an Inkscape template I found online by arlich]

[updated 3 May 2012]
We had to create an A0 template for Android HASS class . Since this is a common problem I searched online for templates to use.

"scientific poster a0 scribus template"
Latek template: http://www.myoutsourcedbrain.com/2008/12/latex-poster-template.html, http://andreas.welcomes-you.com/projects/a0poster/ (A0)

scribus tutorial: http://www.bio.umass.edu/biology/undergraduate/biology-computer-resource-center/documentation/making-a-poster-with-scribus

scribus example outputs: http://nashi.altmuehlnet.de/pipermail/scribus/2007-December/026787.html

inkscape: http://inkscape-forum.andreas-s.net/topic/146897

moar general tips

http://www.nr.no/~strand/poster.html PhD poster HOWTO

http://www.ii.uib.no/~anya/design/ Some oss (gimp, inkscape, scribus) results

http://colinpurrington.com/tips/academic/posterdesign (has good tips, ppt templates)

2005 http://web.pdx.edu/~clarksk/poster.html (good links list / overview different software tools, some dead links, tips)

2007 http://www.flickr.com/groups/postersessions/discuss/72157600205435918/

2.671

Shamelessly ripped from 2.671 materials, because I care more about there being fewer crappy posters in the world:
https://wikis.mit.edu/confluence/display/2DOT671/Posters

MIT/cambridge-related startups&webapps, web 2.0

I keep losing track of these links in the infernal mess of my inbox, so here:

• http://mit.collegecraigs.com/
• http://www.nchoosetwo.com/
  " a service where you can suggest (romantic) matches between your friends.  A friend and I created it last year, and MIT had quite a bit of fun and success with the matches."

HACKNY LINKS

PICS
APP
GITHUB

experiments in education: traiging assignments

Objective
Load system with more credits than manufacturer's spec max classes for optimal performance (A's), input command to pass classes, how long and does the system recover? System should be designed to maximize learning.

In other words, if I take way more credits than I should (based on experimental data from past semesters) for doing well, but only try to pass them instead of doing well in them, will I be able to pass all my classes?

System Design
Consider triaging of class assignments, as described in http://tech.mit.edu/V118/N15/dunn.15c.html.

more fitbit / sleep research

fitbit vs sleep cycle http://www.healthyobsessions.net/2010/12/sleep-tracking-looking-at-fitbit/ ,

1. The tracker is directly on you 2. They only make claims they can back up They want to make sure they can prove their claims, so they don’t try to deduce or diagnose from the data they collect.

Hmm, interesting because I've repeatedly had people tell me that data is interesting but only for a week before people get bored. What is key is making the data "actionable" or providing "suggestions" / expert recommendations.

It may related to casual versus serious user audiences: http://www.graphpaper.com/2010/03-07_for-sale-fitbit-like-new

Fitbit sleep data internet research

Brain/linkdump, got my first set of data out of fitbit properly (didn't realize* you could (a) hold down button for start/stop -> was recording times by hand and inputting manually (b) needed to wear on wristband on non-dominant hand to get accurate(ish) data, not just leave it in my pocket),
*fitbit doesn't come with a manual and the "help" link on the website is pretty tiny so I didn't bother 'till today

This morning 13 Mar 2012

N51 Machine Shop -- persuasive research

2009 http://tech.mit.edu/V129/N31/singapore.html

•  a new kind of university centered on “integrated/holistic technologically centered education,”
•  highlight an approach that is “interdisciplinary with fluid organizational boundaries” and include an emphasis on design and industrial involvement.

Portfolio Apps for Engineers, comparison

It's internship-hunt time for undergrads~
I am pretty terrible at this whole process and don't really have a portfolio setup besides a picture slideshow off of picasa. This seems to not have received positive responses from people like Apple. Oops. ^___^;;



Yea, that's right, I literally emailed them a link to http://orangenarwhals.blogspot.com/p/projects.html. *sulks* I mean, surely the nyancats and hexapods would speak for themselves.

Some of my friends brought up that uhhh maybe this is not very professional and p.s. it looks like it was just thrown together.

This is true of course.

Then I contemplated making a real portfolio. But really, that's not what I want to spend time on right now. This is a common problem, surely there must be webapps for this...

Whoa look there's a list of portfolio apps!
Also, there are wordpress themes for portfolios.


In the end, I think Carbonmade wins out.

FolioHD
I like folio because they integrate videos from youtube/vimeo. Free version:
36 uploads, +mulitple image uploads, +mulitple galleries, +videos from youtube/vimeo.
-resume attachment. Which is fine for me, since I can just include a link to my resume pdf.
The main drawback is lack of theming, so I can't change the default homepage look so that it doesn't look so ugly with just two galleries.
paid features comparison: https://foliohd.com/upgrade

Mostly I like that youtube videos are integrated. Too bad there's not much choice in terms of free styles.
http://nancyouyang.foliohd.com/all

Ravael
I would like ravael (esp. since they limit to 100mb, not to #pics) but I like the image display flash popup thing less, so they lose on theming, but they do allow custom themes (selection of layouts and then you upload a CSS file).
http://raveal.net/p/nancy29888/portfolio
Paid features comparison: https://raveal.net/people/changeplan?execution=e1s1

Krop
Krop would win in terms of the default free aesthetic if they didn't limit me to ten images and no videos :( Also they don't allow multiple-image-at-once upload for free.
http://www.krop.com/nancyouyang/
Free: Basic10 image limit, classic-light style only.
Paid features: https://www.krop.com/profile/creativepro/
(also, look, pretty art: http://www.krop.com/terryfan/)

Carbonmade
Carbonmade wins on account of their user-interface and default free theming. Every change is live. And they allow a bit of homepage customization so that my homepage doesn't look so empty with just two galleries. But I can't integrate videos :(
http://nancyouyang.carbonmade.com/
Paid features comparison:  https://carbonmade.com/signup ($12/month for 50 projects500 images10 videos)
Free: 2 of 5 projects19 of 35 images0 of 0 videos

Magnt
Magnt is more like ... I dunno, some sort of generic aggregated web presence. Does not fulfill my need for a photo/video portfolio of my projects.
http://nancyouyang.magnt.com/

Wix

I tried out Wix too, but don't like it as much. It's for general site-building, hich is nice and amazingly easy to use / customizable, but too much overhead for my dump-pictures strategy.
http://www.wix.com/nouyang/portfolio#!portfolio


Conclusion
Ultimately, I hope to roll my own Flask site, but in the meantime. Back to psetting.

===
notes to self: list of ruby (jekyll) sites + source code
https://github.com/mojombo/jekyll/wiki/Sites
Themequery -- if only the galleries where recursive, hmm.
http://themequery.com/portfolio-template/
Best: http://wpesp.com/

minimalist todo lists

Hope to replace google docs text-based day-level scheduler Cathy Wu introduced me to, e.g. see Metrify todo. I want browser-based (I use desktop+netbook+school computers, I don't use a smartphone or tablet). Google calendar... I don't understand it's paradigm (perhaps it's for outlook-trained users?).

• day  http://tomorrow.do/ --> allows nesting
• timer http://tomatoi.st/ajle --> browser based pomodoro timer
• week http://teuxdeux.com/ --> no nesting. Add times like on gdoc, [3-5pm][
• email http://www.gtasker.com/ --> integrated with gmail, makes me appreciate the todo! but no weekly level overview

I feel that I would be really happy if gtasker and teuxdeux and tomorrow.do got together and had a baby. Or at least talked to each other, if they had an API or something like google scripts I would so totally write email scrapers that would stick things onto teuxdeux, and each day at 6 am teuxdeux things would get transferred to tomorrow.do and each night at 5am the things I pushed off until tomorrow would sync up on teuxdeux...

I also want to script an email reminder every morning, because Gmail is an universal interface I will never forget to login and check. And I want to write a text based interface to everything -- something like Paprika/Workflowy's text-based input system. I guess I'd need to write an intermediate script to scrape event data from my email (all the MIT events are sent out via email and I wish I could train gmail to better recognize events even if I needed to give it a bit of human input, e.g. tell it this email has an event and force it to guess the event data) and input it text-wise.

I wonder if I can scrape these sites to implement a limited version of these features...

/me adds to project todo list... on one of the webapps I just mentioned, but which one? hah.

Speaking of which, I just rediscovered workflowy.com which I was in love with for a while but then forgot all about it. I guess there's something to be said for those pesky once-a-month emails from services I rarely use: at least I don't forget about their existence entirely...

Reference: quora topic
http://www.quora.com/To-Do-List-Applications
Alternative timer (less control, but has audio / alarm):
http://www.focusboosterapp.com/live

to caffeinate or not, as a student

last semester, I began taking caffeine pills, but in a very moderated amount (so probably equivalent to people who drink soft drinks / energy drinks, e.g. 50 to 100mg a day / a quarter to half a pill, aka one cup of coffee, but only on days I reaallly needed it). I wasn't terribly bothered by the stigma of taking pills (I don't like the taste of coffee, I don't particularly like sweet soft/energy drinks, I'm too lazy to keep brewing tea, and I'm a cheap student: it's way cheaper to obtain caffeine in pill form than liquid form). However, I was hesitant due to the stigma of requiring chemicals to help my be productive (e.g. wondering if I was deficient in not being able to simply motivate myself to work harder). Also, I wanted to save my caffeine tolerance for when I really needed it.

This semester, I've decided I don't really care how I get my work done and I have no pride in independence from chemicals. Also, encouragingly:

http://chowhound.chow.com/topics/599101

It's not just the caffeine that needs to clear from your body-- you need to wait for your enzyme levels to return to baseline. Constistent exposure to caffeine will induce the expression of the enzyme that metabolizes caffeine (CYP 1A2) in your liver, so you'll also need to wait until the elevated levels of enzyme decrease in order for your caffeine tolerance to subside. I'm not exactly sure how long that takes (probably a couple of days)

It seems the caffeine tolerance is not lifelong permanent but rather something that can be fixed in a week or even a month.
Interestingly,
http://chowhound.chow.com/topics/808433

For some people with ADD/ADHD caffeine has a calming effect and can even make them drowsy. They seem to be immune to the 'hyper' effects of caffinated substances which only last short-term.

Hmm, I've definitely fallen asleep right after a caffeine dose last semester, and I doubt at the time I had much of a caffeine tolerance but cannot be 100% certain (how fast does tolerance build up? a day? a week?). I don't have ADD though. I think. Perhaps it's partly due to how confident I am in my ability to fall sleep no matter what I try;
http://en.wikipedia.org/wiki/Health_effects_of_caffeine

Studies with placebo conditions have been done to demonstrate that the effects of caffeine depend greatly on the consumer’s expectations

add/drowsiness from caffeine: Backed up by research?
http://www.chicagoreader.com/chicago/how-does-caffeine-affect-nervous-system-health-research/Content?oid=875717

"Caffeine has a tremendously wide variation in action," Regestein admits. "The people who say 'I can drink a cup of coffee right before I go to bed and go right to sleep' aren't lying." Hard biochemical research confirms the fact. Carney describes "one common strain of laboratory mouse, Jackson's Lab's SWR strain, inbred since the 1920s, who is just totally immune to the effect of caffeine: there's no dose that will excite him--not 100 milligrams per kilogram, which would be equivalent to 100 cups of coffee in a human.

After a quick search on google scholar, I don't really see any sources for caffeine effects variability in humans being linked to ADD/ADHD so I will regard it as unsubstantiated rumor.

The plan is that this semester, I'll experiment by dosing either 200 or 400 mg / day (one pill, around 2pm and 10pm) and see if that helps me stay motivated to finish all my psets. Of course, this isn't a sound experiment in any sense, since I believe I am less burnt out, more self-confident, less likely to make excuses and more motivated than the beginning of last semester. (But perhaps I was this on top of things last semester and excited, and things fell from there. I don't remember). This anecdote in particular:
http://www.reddit.com/r/confession/comments/nxdzz/im_not_as_smart_as_i_thought_i_was/

I am an identical twin. Cognitively, my twin and I score the same on tests, and we largely have the same interests (algorithmic analysis, specifically, programming).

In high school, my twin (the paperwork claims twin "A"), spent a lot of time doing his math homework. Like, all of it. Sometimes, all the problems in the chapter, maybe more than once.

Of course, I took advantage of this and did ... none of it.

Fast forward to college, and although we are still performing the same on tests, the reward for his diligence becomes obvious. I struggle through Calculus 1 and 2, combinatorics, linear algebra, and proofs. He picks up a Math degree along with two others at the same time ... a year faster than I complete my degree. He picks up stats, applied math, and a few other esoteric things (like x-ray crystallography) for fun. He does fundamental work in cancer research before he graduates.

Grants open up for his PhD work, and he gets his choice of programs at different schools. I have to pay my own way for a Master's degree. He gets kicked from the GCC mailing list for posting too many bugs. I almost get kicked from graduate school because I can't keep up with the lab.

Suffice it to say, it all came down to his early time and effort actually doing his homework. Seriously. The worst part is, after putting in several months at the beginning of slogging through his Maths homework, he could do all of the problems much faster than I. So, while I toiled on the minimum set, he could hammer out the homework in minutes and be that much farther ahead.

In addition, working with very motivated people this winter break has helped lift me from my skate-through-life attitude toward not completing assignments.

I hope the breaks (weekends possibly, spring break, summer break) will give my body time to reset itself. I've also allowed for ramping for each semester -- 600mg/day for senior fall, 800mg/day for senior spring -- in case caffeine is effective but I start developing a permanent tolerance to it. I really think I'll be able to stay on <400mg/day though.

Dosing research / Safety Check
http://www.caffeinedependence.org/caffeine_dependence.html#tolerance

 In caffeine non-users or intermittent users, low dietary doses of caffeine (20-200 mg) generally produce positive mood effects such as increased well-being, happiness, energetic arousal, alertness, and sociability. Among daily caffeine consumers, much of the positive mood effect experienced with consumption of caffeine in the morning after overnight abstinence is due to suppression of a low grade withdrawal symptoms such as sleepiness and lethargy (see section on Caffeine Withdrawal). Large caffeine doses (200 mg or greater) may produce negative mood effects. Although generally mild and brief, these effects include increased anxiety, nervousness, jitteriness, and upset stomach. However, individual differences in sensitivity and tolerance affect the severity and likelihood of experiencing negative effects [...] high dietary doses of caffeine (200 mg or more) increase anxiety ratings and induce panic attacks in the general population. Individuals with panic and anxiety disorders are especially sensitive to the effects of caffeine [...]  High doses of caffeine (750 to 1200 mg/day spread throughout the day) administered daily, have been shown to produce "complete" tolerance (i.e., caffeine effects are no longer different from baseline or placebo) to some, but not all of the effects of caffeine. 

http://en.wikipedia.org/wiki/Health_effects_of_caffeine

For the rest of the general population of healthy adults, Health Canada advises a daily intake of no more than 400 mg.^[46] 

Lethal dose research: http://en.wikipedia.org/wiki/Health_effects_of_caffeine
150mg / kg, 160lbs = 70kg (yes, I'm becoming more and more overweight), lethal dose = 10.5 grams. I should be safely under this.

Additional research
http://www.quora.com/How-bad-is-it-to-drink-800-to-1-500-mg-of-caffeine-daily

 the average US coffee consumption is about two cups per day, which is the equivalent of approximately 280 mg of caffeine daily. [1]

Seems on the high side to me o.o I never realized we consume so much coffee.

http://www.quora.com/Are-there-any-permanent-effects-of-caffeine-intake-on-brain-structure-or-function

For adults consuming moderate amounts of coffee (3–4 cups/d providing 300–400 mg/d of caffeine), there is little evidence of health risks and some evidence of health benefits.

Okay, so my caffeine consumption isn't too crazy then, even if right now I feel amazingly less sluggish / craving sleep than normal.

Tolerance
Lifehacker draws on dated studies:
http://lifehacker.com/5585217/what-caffeine-actually-does-to-your-brain

A 1995 study suggests that humans become tolerant to their daily dose of caffeine—whether a single soda or a serious espresso habit—somewhere between a week and 12 days. And that tolerance is pretty strong. [...] caffeine withdrawal very quickly, anywhere from 12 to 24 hours after your last use [...] you'll be over it in around 10 days

Conclusion
200 [to 400] mg / weekday: 2pm, [10pm]. None on weekends.
200mg seems like a sufficiently high intake compared to my normal levels (<50 mg/day all winter break / most of my life) that I should be able to determine if caffeine will help me be more productive, or if really it comes down to a self-motivation problem no chemical can fix.

~~~~~~
to figure out:
can I measure the actual caffeine / metabolites in my blood levels for meche instrumentation 2.671 class? seems like the international sports people relied on some sort of urine test.


~~
random copy paste from failbook
[7 Mar 2012]
TIL about cough drops and caffeine: "Menthol did approximately double the time to maximal caffeine serum concentrations, however, suggesting that menthol delays the absorption of caffeine." Hmm. Sounds like a possible stopgap for accidental extra caffeine. Except this was wrt 100mg of menthol and 200mg of caffeine and the average cough drop ~8mg of menthol.(http://www.pharmacytimes.com/publications/issue/2010/May2010/DrugInteractions-0510)

[7 Mar 2012]
hmm. caffeine in pill form, at 12g ~$0.23 / g (price seems to have gone up since october 2011). caffeine in powder form, at .5kg ~$0.06 /g. >> to investigate: shelf-life and personal consumption rate, but I bet it's not worth it.

[14 Dec 2011]

Nancy R. Ouyang I've input about as much caffeine over the last 48 hours as previously over the entire semester. CHEMICALS, WHY DO YOU FAIL AT KEEPING ME AWAKE ;____;
December 14, 2011 at 5:16am · Like

A strongly recommend making your plane to signal spacing much wider
December 14, 2011 at 5:22am · Like

B Appendix 1 to caffeine tips:
* Caffeine can only do so much to override sleepiness before fatigue takes over and kicks you in the face.
* By "face" I mean consciousness.
* The solution is NOT stronger stimulants. Remember, awake != functional.
December 14, 2011 at 3:43pm · Like

Nancy R. Ouyang oh, no worries, i get demotivated and fall asleep long before i stop being functional. derp! :D but thanks.


Nancy R. Ouyang
ugh, 5 different labs to split my time between. 6.131 (9-23, and the one that edges me into overstressed-land), 2.008 (7-16), mas.863 (anytime), MITERS (anytime), UROP (set times). I think it's time to investigate caffeine pills so I can tip the scales in favor of building things 24/7 (vs indulging my love of napping) because now I actually have infinity projects to work on.
Like ·  · October 25, 2011 at 12:20am

B TIPS:

* Cut pills in two. 200mg is a fairly high amount for one dose and you'll feel better (read: less jitter) with an extended dose over time.
* Combine with aspirin or an NSAID (like ibuprofen) for slight synergistic effects.
* LD50 is roughly 150mg/kg body mass. That said, watch your dosage. Heart palpitations suck.
* Being awake != being functional.
* Amazon Prime is your best friend for instant gratification.
October 25, 2011 at 1:54am · Like ·  1