April 2009

Integrals fall from the sky

Back when we were slogging through Calculus, my friends and I used to complain that there was no practical use to what we were learning.  What were the chances, for instance, of encountering a Taylor expansion in a dark alleyway?  Or of an integral falling from the sky?  Except in my world, of course, where this happens all the time.

Suppose, just for instance, that you are designing an auditory stimulus with certain statistical properties and you want to know what the overall power in the stimulus will be.  Or that you make a new stimulus composed of lots of independently fluctuating signals at different frequencies and you want to know what happens when you make the individual signals vary more (answer: the mean power across frequencies increases).  Or that you can measure a signal after it’s been slowed down by an odor delivery device, or by an odor delivery device plus a neuron, and you want to figure out how much of the slowing is due to the neuron.  I swear those integrals are everywhere.

Around year 2 of graduate school I got very depressed about the fact that I was never going to be the sort of person who could look an integral in the eye and not blanche.  The nadir was the Computation Neuroscience course at Woods Hole, where well-meaning theorists liked to comment that those of us running experiments had no idea what our data meant and Haim Sompolinsky would pop by in the evenings to drop off fun problems and to glower menacingly when we couldn’t solve them on the spot.

Towards the end of grad school I came to terms with the fact that—although I’d never be the kind of person who could solve the integral myself— I could at least be the sort of person who recognized when there was an integral there to solve.  Thanks to Matlab (and many years in a windowless closet) I have a pretty good visual intuition for a number of mathematical concepts.  I am—if not at home in frequency domain—at least able to read the road signs and ask for directions.  And when an integral falls from the sky I know to call my Dad.

The beauty of asking a mathematician for help is that they send you the answer in Tex, which makes it look deeply professional and well thought-out, and gives me the lovely impression that what I’m doing must be Real Science.  Either that, or on napkins, which is if anything even more bad-ass.

I know I don’t post about science much, even though it’s up there with art and sewing on the masthead and is ostensibly what I do all day.  But I just wanted to say thanks, Dad.  You’re the best.  And I’ve still got those napkins from grad school.


Comments (1)


It wouldn’t be an experiment if it worked the first time

From a recent conversion with my labmate Quentin:

Quentin:  Weren’t you doing some RNAi experiments?  Whatever happened to those?

Me:  They worked.

Quentin:  They worked?

Me: Yeah.  How awesome is that?

Quentin:  That’s so boring.

Fortunately, sewing hasn’t been so boring.  Inspired by Carolyn’s excellent series on designing her own patterns from a TNT (tried and true) dress, I decided to have a stab at creating my own top pattern.  I really liked the shawl-collar top on the McCall’s knit dress (5752), and I wondered if I could make a simpler version of it—without the ruched midriff— by combining the dress bodice with my favorite T-shirt pattern (Simplicity 3634).  

To start, I copied the t-shirt front onto a folded piece of tracing paper, using the V-neckline rather than my usual scoop-neckline.

Then I laid the dress bodice front on top, lined up the shoulders and side as best I could and traced the sleeves from the bodice.

I cut out the pattern piece and drew a line across the front for the wrap, then cut out this piece.

I copied this shape onto another piece of tracing paper and added the “shawl,” modeling it on the one in the bodice top.  I also added sleeves to the back piece:

When I cut this out of fabric and held it up to myself I could see that (a) the base of the wrap was too high, and (b) the shawl didn’t have enough body to it.  Also the curved line I cut for the wrap didn’t really work for the shawl.  Then it occurred to me that I could make the shawl out of a separate piece of fabric.  So, I cut the shawl parts off the fabric pieces, and cut a separate shawl piece.  I used the cowl pattern piece from my t-shirt pattern for the width (11″) and calculated the length based on the length of the back neckline plus the two sides of the wrap (27″).  I assembled the top without the wrap, leaving holes in the side seams where the shawl/wrap would fit in.  Then I folded the shawl piece in half lengthwise and pinned it all the way around the top:

I stretched it a little as it went so it would pull in and stay closed, so I ended up with about 2″ too much on either end.  This was fine as I just cut the ends off after I sewed the shawl down.  The shawl doubles over on itself so you end up with four layers of fabric.

Here’s the top as it stands.  I think it makes me look like a samurai working for UPS.  I can’t tell if the style would work in another color or if I should scrap this idea entirely.  I think if I were to make this again I would:

  1. Use regular cap sleeves instead of the extended shoulders
  2. Make the neckline a little narrower
  3. Possibly reduce the wrap to something more like this (don’t know what its called—the line drawing is from the Jalie website): 


Comments (1)


Shades of gray, part II

A gray jacket

Work on the vogue jacket is proceeding s.l.o.w.l.y.  Partly this is because I am actually taking my time, clipping and finishing my seams, for example, so the whole thing does unravel after a week.  Partly it’s because it has been hard to fit.  I actually threw the muslin in the trash at one point, although I did dig it out and continue working on it.  Here are the final muslin pieces laid over the original pattern pieces.  You can see I took a big curved chunk out of the back to account for the fact that my spine is less than straight.   The bust darts are also still too high.  I made them a bit shorter so its not so noticeable but next time I will have to move them down.

The shell is basically done, as you can see above.  I also made a version of the lining:

But now I have to make important stylistic decisions, such as Should it have pockets? and What kind of pockets should they be? and Might the lining be a tad too stripey?

The answer to the first question is of COURSE it should have pockets.  What kind of a roo’s jacket would it be without pockets?  But the second one is more difficult.  I’ve thought about patch pockets:

Although I think it would also look nice with welt pockets:

The disadvantage of welt pockets is that I have no idea how to make them.  Not that this has stopped me before.  I also wonder if it would look better with a solid color lining:

What do you think?




Comments (2)


What I’m working on—shades of gray edition

 A gray painting

Since the show in December I haven’t been painting too much.  This is partly because I made the unfortunate discovery of online fabric stores, and partly because I didn’t know what I wanted to paint.  Or rather, I did know: I wanted to get more texture in my paintings, I wanted to paint things that worked both as landscapes and as abstractions, and I wanted to paint scenes that were dimly lit but still luminous, where the color and light arise out of many subtly different grays  But I didn’t have good source material to work from—until it started raining a few weeks ago.

Here is the photo I worked from.  Its the view from outside my building at work and I’m trying to snap it quickly before the bus in the right hand corner leaves without me.

I tried to keep the initial pencil sketch loose—so the painting would read as large areas of gray and small spots of color, rather than as definite buildings.

For the first layers I used only a large brush and blocked in the main tones and shapes.  I used only four brushes on the whole painting—partly to cut down on the number I had to clean, and partly to force myself to use a “broad distribution of spatial frequencies.”

In the next stage I played up the differences in hue between the adjacent buildings and deepened the shadows relative to the photo.  

I liked how the building at the top center worked at this point, but the yellowish building looked pretty flat.  And the area to the right where the street disappears didn’t “sparkle” the way I wanted it to.  After looking at the photo I took of the painting I realized that the yellow building needed to be warmer—more peach than mustard.

When I went back into the painting I warmed up both the yellow buildings and the building at the far right by glazing over them with a light peach color.  I also added a warm gray (indian yellow mixed into quinicridone rose and prussian blue).  You can see it on the side of the bus and on the foreground of the street.  This color looked pretty ugly on the palette but I think it helped the painting a lot,.  It gives the bits of orange light on the building something to relate to and makes them look like light and not just bits of paint.


Comments (0)


Dress like a nerd

There’s a widget on this blog that lets me see what search terms people are using to find it. This morning, apparently, someone linked to my blog by searching for the term “dress like a nerd.” I can’t tell you how pleased this made me.

Which got me thinking about how nerds actually dress. Having observed nerds closely in their natural habitat I can verify that the days of lab coats and pocket protectors are long gone. The emphasis in the lab is on comfort, however achieved. For example, when I came to interview for my current job, my boss was wearing her pajamas. At least, I think they were her pajamas. They might have been her “dress sweats.” The standard lab uniform, though, is jeans, a t-shirt, and a polar fleece. The t-shirt may be either ironic (“Stand back, I’m going to do SCIENCE!), or have come from a conference, vendor, or summer course—the latter kinds having the advantage of being free.

There’s something wonderful, of course, about scientists’ complete disregard for presentability, but after a while some of us start to want to…well…look like grown ups. At least occasionally. This is easier said than done. For one thing, there are functional reasons for the standard lab uniform (S.L.U.): Labwear needs to be flexible and concealing— there’s a good chance on any given day that you’ll have to crawl around behind or under something to move some wires or adjust some tubing. It needs pockets to store your ID card and keys. And it needs to keep you warm as the temperature in the average lab is only slightly higher than the temperature in the average refrigerator.

There’s also the power of conformity. Those who break the lab dress code by wearing, say, a nice pair of slacks, a sweater vest, or—God forbid—a dress, are likely to be asked where they think they are going and severely teased. Not being immune to this myself I have evolved my own take on the S.L.U.: wide leg jeans, fitted solid-color t-shirt, and a zip-up cardigan. This allows me to blend in adequately while feeling marginally more put-together than I did at age 13. After many years I have gotten to the point where I can occasionally pull off wearing a skirt. (So long as my sweater has sufficient pockets. Nobody, but nobody, carries a purse to lab.) But I still can’t bring myself to feel comfortable in a dress. Which is why it’s unfortunate that I made plans to sew a whole bunch of dresses this summer:

Perhaps they can be adapted into tops? Otherwise I guess I’ll be making many things for my sister (the top center one is for her). What do you think, fellow nerdlings? What is appropriate to wear to lab?


Comments (5)


Heavy Table

My good friend Jim, food writer extraordinaire, has a new site dedicated to food of the Upper Midwest. Check it out.


Comments (0)


Beautiful soup of the eeeeeevening

After many months of resistance I finally caved and bought a second kind of Korean bean paste. The seasoned soybean paste, or Sagyegeol Ssamjang, comes in a green container, is the secret to delicious bi bim bop and mapo tofu, and has been a staple in our kitchen for over a year. The hot pepper soybean paste comes in a red container and looked intimidatingly spicy—until now!

After consulting the internet I came up with the following recipe for spicy fish soup:

  1. Boil water. Put fish is water.
  2. Take fish out of water, add bean paste.
  3. Optionally: add ginger, garlic, soy sauce, sherry, hot pepper flakes, tofu, veggies of your choice.
  4. Take fish off bones, put fish meat back in soup.
  5. Sprinkle green onions on top.

Make sure your bed is nearby and made up so you can collapse moments after finishing your soup.


Comments (1)


Might as well be spring

Spring used to be my least favorite season.  Growing up in Wisconsin, I though of March through mid-May as just the extended part of winter.  The trees were bare, there were lumps of dirty snow and partially decomposed leaves on the ground, and just when it looked like it was really going to warm up, it would snow again.

I started to change my mind on our honeymoon trip to New Zealand.  That was the first time I realized that when the trees first put out their leaves, they have a little of their fall color.  The spring colors in New Zealand were incredible, dark branches and bright greens against moody skies in a dozen different grays.  It was these colors that inspired me to start oil painting again.  

Then we moved to Boston—in January—and by March both David and I were convinced that it was going to be cold for the next six years.  Walking around town at the end of March though, I finally noticed the city getting ready for spring.  San Francisco has a desert spring.  It rains in January, everything blooms as fast as possible, and it’s dry again by March.  But here the plants can take their time.  The geese come back in February.  The trees start budding in early March.  It reminded me of an orchestra warming up.  Mostly cacophony, a lot of scratchy notes, and then every once in a while a fully formed melody would burst out from a soloist who couldn’t wait for the concert to start.

So, it’s spring again.  The skies are lovely and rainy and overcast.  And I’ve been out with David’s camera taking pictures of all the beautiful grays:

I’m going to use these as source material for more paintings in the “Statistics of Natural Images” series.  (Look at all those large areas of muted grays!  And the bright spots of saturated color!)  That’s my excuse for taking shaky pictures.  Also that I’m a bad granddaughter.  At least I didn’t use a flash.

I think I mentioned before that I’ve also never been crazy about traditional spring colors for clothing.  So this year I am trying to do spring style à la Roo by sewing floral prints in rooish gray:

And in tones of the 70’s:

I’m also working on this stripey gray and pink jacket (the pink stripes are for the lining).  I swear this collar fit the neckline when I made the muslin.  Now it’s about 2 seam allowances too long.

So long as it keeps raining, I should be able to finish it.


Comments (3)