this week i

released the 2013 employer health benefits survey.  well, nirmita and matthew did.  and gary answered reporters' questions like a champ for an hour.

posted code to analyze the major american longitudinal teenager health study, littered with references to ancient greece.

cleaned up, moved out.  living in the shrine to my youth for a little while.

consider publication bias a part of our dna.  for example: name someone who attempted to cross the mid-atlantic before colombus.  i can't either.  remembering failures goes against something primal.

drove up to new jersey for the food.  this is joe, my father's best friend.  pictured here with blueberry pie.

ate at a days inndian restaurant and other favorites.

ran a solid five miles across brigantine island, then ate with my father, joe's neighbor nick, joe's partner linda, joe himself, my cousin lynne, and my aunt rosemarie.  you wouldn't believe how many puppets i had to go through to get my father to smile for the camera.

spent a lazy sunday morning writing algorithms.  no help from these two.

8/29

this week i

figured out the share of health insurance premiums subsidized by the affordable care act in the self-purchase market.

watched my old professor discuss the compromise of reproducibility (at 7:05).  with their deterministic statistical machine idea, roger and jeff are way ahead of both my work and the situation on the ground.  they're searching for the statistical equivalent of einstein's unified field theory while i'm over here tying metal keys to kites in thunderstorms just to see what happens.

crossed the bay bridge.

camped at maryland's finest.

met the locals.

have a camera in my pocket, thanks.  although i was happy to see howard and misty.

snapped a tent selfie.

shut down for forty-eight hours.

suspended the rules.

look forward to next year's double-rainbow-themed birthday cards.

received twenty-two tuples.  my first gift of obfuscation.

8/22

this week i

published code to look at, well, um, every doctor in the united states.

received the official endorsement of the brazilian census bureau.

jotted down my idea for a data set with three hundred million synthetic americans.  i'll get to it as soon as i tire of asdfree.com.  so possibly never.

patiently waited for doctor erin trish to finish that glass of water so we could celebrate her defense with this bottle of wine.

had a long conversation with the uk data service about how to automate the login.  they're going to make it easier.  and legal.  six to eight months.

got monetdb export coaching from hannes - copy select * from table into foo.csv using delimiters ',' null as '';  also ought to use fetch() asap.

scrubbed mold off the ceiling.

traded two burritos, two six packs for a freshly-painted bathroom before i move out.

 

8/15

this week i

packed one extra in the suit-case.

met hannes in-person, in-baltimore.

tried on the best of baltimore party, themed for..

prohibition re-enactors

mad scientists

henry ford look-alikes

mustache swappers

once-in-a-while troubadours

hunter-gatherers

previously hungry people

cosmetic surgery center of maryland success stories

village people reunion tours

not pictured: post-party surprise free cupcake for random strangers distributionists.

climbed up to nine for the workday.

ate thanks to - and let's hope this sticks - grillmaster siva.

sailed to key bridge.

hung out with one of the rare folks who speak my language.  he scolded me for my continued use of `sample` in column names, "it's a freakin' sql reserved word."  who else can i talk about this stuff with?  nobody, that's who.

rest my case.

8/8

this week i

published the code you need to examine the youth risk behavior surveillance system.  if i don't do it, somebody else will.

drank a bottle of wine with danielle.

became a landlord.

don't see the utility yet, but maybe that's the point.  shooting in the dark - how science works.

went by anthony d apostrophe capital a mico in high school.  name changes: yet another thing me and don draper have in common.

snared our first help mailing list question.

8/1