Adam Litterman's Personal Homepage:

What's new?

Quite a bit!

These are some of the web applications I've built (last updated 10/24/09).

Correlations between U.S. state-by-state data sets

A relatively simple and nerdy but somewhat interesting web application I built over the summer and finished right as school was starting that I've been meaning to put a link up to. There's a link below to some of the continuous mapping of two-dimensional variables I find interesting, and this web app basically takes it to its logical extent. It uses a flash based piece of software that shows all 50 states in various colors and takes any two variables from a list of 20 or so data sets that I've compiled for all the states that score everything from murder rates to the price of health insurance to population density to statehouse political composition to geographic area etc. etc. and allows you to pick any two variables and look at 1) a "heat" map that shows the various states and their relative intensity on a given variable relative to the other states and when you click a button under the map 2) a graph of the data showing where the states fall on the X and Y axes (which shows you how their color was picked) as well as 3) a trend line that shows a simple linear best fit line and finally 4) that best fit line's slope, correlation coefficient (r) and coefficient of determination (r².) This final value, the coefficient of determination, is interpreted as "the proportion of variation in the independent variable accounted for by the dependent variable," and if you play with the thing you'll find some weird correlations and some absence of correlation that may or may not surprise you.

dynamicDebt.org

Another web application I built that's much simpler in scope than the baseball one (only one you page see, total, and the whole application can be used in about 20 seconds) that addresses a subject that I've talked about on the web site before, just how bad the debt is that this generation of politicians is leaving for my generation and younger generations to deal with. Thanks guys! The web site uses some basic AJAX and PHP (see below) to allow you to see the consequences to our deficit and debt picture if you input into economic forecasts more realistic estimates of what GDP growth will be year by year from 2010 to 2019... Seriously, the official government estimate from the Congressional Budget Office is like, not only does the recession stop next year and turn into 2 percent and change growth, but in 2011 our economy grows at a scalding annual rate of 5%, and then turns into a sustained decade long boom... and that glass is 7/8 full outlook on life still leaves us $14 trillion in the hole, holding a national debt equal to about 70 cents of every dollar of wealth generated by Americans in a year. Adjusting the GDP growth estimates to a more pessimistic (and in my mind, more realistic) situation... the recession dragging on for a year or two longer with less robust recovery for the rest of the decade... is like watching a train wreck you can see coming: You Just Can't Help But Look.

MyMLBSim.com

The biggest project that I've been working on is one that I haven't done a great job of branding... The domain of the whole web-site is baseballmarkovchain.org because the idea was to promote a baseball statistics simulation program built on a Markov Chain. Then the project sprawled way beyond what I imagined it would be and the fact that the project used a Markov Chain as its mathematical model core became less important than the fact that MyMLBSim.com has become a multi-faceted baseball statistics simulator that lets a user set up custom models of any or all of the 30 MLB baseball teams and then simulate how changes from predicted baseline player performance (a player over-performing or under-performing in a certain way, for instance, by hitting more home runs than expected, or hitting in a different spot in the lineup) effects both: 1) a single team's play (how many runs it will score, how many runs it will allow, how its offense will score those runs, how they would expect to do over a full season and 2) league-wide simulations such as how likely certain teams are to win the actual remaining games on the schedule or what the likely final standings are and what any given team's chances of winning the division, the wild card, or making the playoffs are.

So the progress of this sprawling web app that's now tens of thousands of lines of code in HTML and Javascript to create both what's typically called AJAX in the sense that some of the pages get data from calculations done on the server while you stay on the same page, which allows for an interactive experience when combined with the mix of Javascript and HTML known as Dynamic HTML where interactive web pages don't require static forms to be filled out and submitted... there are windows that you can drag around and close, buttons you can push to change things on the page, sliders and drag and drop tokens to get user input, and so forth. PHP is used to do almost all of the of the heavy lifting of the computations and data fetching from other web sites that computes things like who's on what team, who's won what games, what the schedule is, what a player's stats and predicted baseline performance are, drawing the pictures for simulations with visual output, etc. Finally, the data of what users are signed up, what teams they've configured, what they predicted about various players and how they arranged them in a lineup, what teams' records and scheduels are, etc. are done with MySQL, which is a programming language which allows you to create and use databases--allowing me to automatically keep thousands upon thousands of pieces of data that make the site whir organized and accessible. While the three are more or less glued together, unfortunately, Internet Exploder has this thing about not following the standardized specifications for style sheets (how modern web pages are both dressed up with colors and fonts and so forth but also how they're arranged, formatted amd organized) and Javascript. This means that for now the application cannot even be considered an alpha version (which would be a mostly working version that's known to mostly work but also known to have a fairly significant number of minor bugs.) That's because unless you're using Firefox or Opera or Safari (which is only about 30-40% of the internet population) the program simply won't work. So we're at like version 0.0.9, once I debug the Javascript and make the site work in IE I'll declare it an alpha and reach version 0.1.0, and start going up through 0.1.1 to 0.2.4 until I reach a beta version (which is something like version 0.99, i.e., it's almost completely working) which will hopefully be before opening day 2010.

Blog and other Junk

While it's now getting to be about half myMLBSim.com updates and half observations and political blog entries, my personal blog will both document the progress of the baseball web application and contain other ballyhoo that I don't have time to make specific web pages for. Speaking of which... here is a smattering of stuff I added over the past several months...

Recently I've been interested in showing sociological data on maps as continuous data instead of binned data as is typical which led to me developing some techniques in Photoshop scripting for the purpose of continuous 2-dimensional mapping of social data. This is something I'm interested in, specifically when one dimension of the social data are murder rates in various states and the other are some sociological factors--levels of gun control, levels of poverty, diversity--that might have an effect on murder rates. You'll have to take a look to see the results. Other stuff to look at includes some maps I generated using data about recent job loss and housing foreclosures as well as trying to show how virtually all of the gun-related murders in this country are committed on a tiny fraction of its territory. Rounding out this theme of displaying information visually in unique ways... I'm a big Twins fan, and I created a bunch of "homunculi" representing the 2009 starting batting order. For more info and to see the blog post on Battle Your tail Off that I generated it for, check it out.

About me and some vaguely academic stuff

This is my personal web-site; while there is some academic/intellectual material on this website, for information about my academic career please see my my official University of Minnesota web-page.

Go Gophers!

I am a Ph.D. candidate at the University of Minneota's doctoral program in Microbiology, Immunology and Cancer Biology. I recently moved out of a shipping crate in New York City (or an apartment, I wasn't sure which) and am settling into a lovely new home in Nordeast Minneapolis. live across from these Grain Elevators, and honestly they're one of my favorite things about this particular location (relative to other locations in this neighborhood). Kids play pickup baseball games in the field next to them in the summer and me and my cousin go out there with a bucket full of tennis balls and throw each other batting practice using the massiveness of the grain elevators as a backstop from having to shag balls you really get a hold of. As the summer winds down and the brief fall begins and winter closes in, I'm looking forward to my first Minnesota winter, and to getting back into the swing of school...I'm trying to finish ASAP so that everyone will HAVE to call me "Doctor Litterman," and I will correct you if you don't address me as such.

If you want to follow me on Twitter you can do so here. Oh wait, I'm not Shaquille O'Neal, I forgot. (Hint: Google!) As for some academic links, you can check out an abstract that got included at the book at ASCO last year (That'the American Society for Clinical Oncologists.) I sent a new abstract this year that was presented as both a poster and a power point by my old P.I. (Primary Investigator, i.e. my old boss) to a very enthusiastic response. The data which is described therein is going to be incorporated into a paper that'll be submitted to a relatively high impact journal for Dr. Polsky's lab due to the novelty of the finding, where we showed that within a single tumor a common mutation in a protein that is a melanoma drug target can be present in some parts of the tumor and absent int others. This raises several interesting questions for the Polsky lab and other researchers to pursue about whether this mutation is either necessary or sufficient for causing invasion and metastasis, or even malignancy at all. It's not clear where the paper will be submitted yet and the Polsky group at NYU is hard at work on it as I type this I'm sure (well, figuratively, it's Sunday evening) so I will post a link to that when I learn more. We sent similar abstracts to the World Melanoma 2009 conference in Vienna where Dr. Polsky also spoke. For a completely unorganized look at some of the stuff I generated for Dr. Polsky's group (at least as long as the data is not published in a proprietary journal or I'm told that this shouldn't be put out on the web) check out this folder for miscellaneous snippets of what goes into making a scientific poster--don't worry... this poster has been seen and publicly posted, so it's not like you're getting some magical access to insider information. That would be the pictures from when I worked with monkeys...