Thursday, July 23, 2015

Week 7: Kinks

That title is totally click-bait. But I swear it's relevant!

I finally hit a massive obstacle in my project earlier this week. One of the plots I generated had a pretty nasty kink in it where it shouldn't have had one. This lead to a number of days spent under the hood of my code debugging, re-writing subroutines, checking types, and making sure I didn't exceed the 132 character/line limit (thanks for nothing, Fortran 90).

Today, I finally had a break through. The plot I was generating featured points whose y-values came from an integrand. The integrand is chock full of... interesting quantities that are computational expensive to calculate. The two most complex are the external potential and the structure factor, both of which come from a program that, for all intents and purposes, is a black box to me.

None of my other plots had any issues, so from the outset I was suspicious that the issue was in my code. Still, I started from the top of my integration routine and worked all the way down until it had to be something in my integrand itself that was misbehaving. I checked the fermi-dirac function I was using, the physical parameters that gave the integral correct units, the fineness of my grid, and even how many digits of pi my code was using. Finally I came to the potential and the structure factor. It had to be one or the other. Removing each separately showed that there was an unusual jump in the potential where my kink was. Victory.

The win is hollow though. I still have no idea why the potential is misbehaving and although it looks like that's the issue, I can't really be sure. I just hope that whatever the issue is, it won't have affected the rest of our results.

I can't believe that it's already the end of week 7. Before I know it, I'll be back at sea level and won't need a cryptocard to check my email.

Thursday, July 16, 2015

Week 6: Descent

Today at work, it became really clear to me that the summer is almost over.

I've alluded to our progress in a few of my other entries, but Dan and I are truly at the production stage of our project now. There are a few tweaks and optimizations being done, but most time is being spent crushing the Moonlight data cluster with thousands (literally) of calculations all at once. I sure hope the HPC guys and gals don't completely hate us.

The time one has while jobs are running is a very slippery thing. It's easy to get distracted by an article or a book chapter or (honestly, more commonly), a funny picture of a cat. I'm also not used to having downtime while calculations are running because I haven't needed any HPC for any of my code in grad school up until this point. Sure, I ran stuff in undergraduate, but that was eons ago and I had classes + homework to keep my occupied.

One of the best aspects of having free time at work is that I get to write. Fleshing out the story of our results is far and away the most fun part of doing science in my mind, and goodness me do we have a cool story to tell. I'm not entirely sure which parts of it I'm allowed to talk about and which I'm not, so I'll have to be vague, but the true star of our project will be a set of data that totally and completely surprised us. We predicted that one of our models would behave a certain way and it did the opposite. There's a rationalization we're happy with,  and the result is consistent with our other data, but I love a surprise.

Results that confound and befuddle remind me why I like science in the first place; until you set up an experiment and parse the data, you can't be sure of anything.

Sorry that this entry didn't feature any mortal danger or cattle.

Thursday, July 9, 2015

Week 5: From Pecos, with Love

Last weekend was one of the most interesting 4th of July weekends I've ever had.

Because the 4th landed on a Saturday, we had work off on Friday. I was planning on going into Santa Fe and just relaxing all day when, at about 10:30 pm Thursday night, I got a call from my friend L. Turned out she'd planned on going backpacking with our coworkers G and BC over the holiday and asked if I wanted to go.

A few things worth mentioning at this point; I like the great outdoors and have been hiking loads of times. I fancy myself as reasonably in shape and figured I'd be able to handle ~20 miles of hiking over three days, even if it featured over 3,000 ft of climbing. That all said, I didn't have any of the equipment required for an undertaking like this. Still, L promised that she had enough bits and bobs to cover for me, so I figured what the hell. Why not. I'll go into the Wilderness with L, an avid backpacker, and two other wannabe-adventurers who, like me, had never gone backpacking before.

What could possibly go wrong?

The first day, went splendidly. It took us roughly 3 hours to hike the 6 or so miles to our first campsite. I was surprised how light the 30lb pack on my back felt. It was a cool feeling knowing that the four of us had enough supplies for us all to be completely isolated for an entire weekend. We set up camp near a beautiful mountain lake, summited the peak near Pecos Baldy Lake, ate some s'mores, and relaxed. Other than an overly curious elk that wandered into our camp right as we were going to go to bed, the first day went off without a hitch and I was rather smitten with this whole backpacking thing.

The view from the top of the baldy was pretty awesome.

Saturday continued the good vibes. We hiked another 6 or so miles into the Wilderness and arrived at yet another mountain lake. Along the way, we had to walk along the Trailriders Wall, a beautiful ridge that runs between two heavily forested areas and is about 10,000 feet up. There is essentially nothing up there, so you can see for miles in all directions. It was breathtakingly gorgeous.

Seriously. Miles.

After traversing the wall, we arrived at a mountain pass that was occupied by a herd of bighorn sheep. Keen not to upset them, we stopped for lunch and let a gap open between us and them so as to avoid incident. The rest of the hike was fairly wet, as there's been a lot of rain this season, but we made it to the second lake just in time to start hearing thunder. At this point, we weren't sure whether we should just camp out near the lake, or try and take the return loop back towards the car. The thunder made staying by an elevated and exposed body of water far from enticing, so we opted to plunge back into the forest and began making our way back towards the trailhead.

Thanks to the wet summer we've been having, the loop back to the car was simply gone. In its place was a stream which made it impossible for us to loop to the car. We instead decided to hike back to Pecos Baldy Lake, camp there one more night, and then head back on Sunday as planned. It would be a longer hike, but at least we knew the area around the lake.

At this point, we weren't really hearing thunder anymore, the trail was mostly downhill, and the bighorn sheep were gone! Looked like smooth sailing to the campsite. Then we got back to Trailriders Wall and me oh my did the you-know-what hit the fan.

After climbing back up to the top of the wall, we had only covered a few hundred feet when we started to hear thunder again. But this time we saw the source of the noise: in front of us was a sky completely covered in very, very angry looking clouds. Some of the clouds furthest from us were definitely spitting out rain, and even more were splitting air with a few hundred thousand volts of electricity. Fun.

At this point, we were so far along the wall that we weren't sure what to do. Turning back and returning to the forest for a quick tent-pitching seemed difficult, since there was no cover back the way we came. If we continued forward, there were a few small clusters of pine trees that could at least lure the lightning away from us. We figured pressing on towards the other campsite was our best option.

Then a lightning bolt struck the ground ~300 ft away from us.

I don't actually remember throwing my pack off, but before I knew it, a lightened me and my three friends were full-on sprinting towards the pine trees for cover. We huddled there for a while, begging the storm to relent. It got worse before it got better. The thunder was incessant and we became drenched. There was hail, too.

Luckily for us, the storm did capitulate and we were greeted by a happy rainbow and sunny skies. We decided that another night of camping would be as welcomed as another storm, so we rushed back down the trail, hoping to get to the car before the sun went down. On the way back, we ran into more silly elk, a herd of overly-friendly-to-the-point-of-being-unnerving cows, and heard some coyotes howling as we plunged into the final forest we had to pass, but incredibly, we made it back to the car.

Lots of people say that the Wilderness changed them, and as much as I love defying clich├ęs, I feel the same way. Sure, it was terrifying at times, but there's something special about being that far out in the middle of nowhere. The break from office life was well-appreciated, and now I can't wait to get back to SoCal so I can check out the San Mateo Wilderness!

Smiling atop Trailriders Wall. Like a moron.

Thursday, July 2, 2015

Week 4: Thundercat

In case you don't know, Thundercat is this groovy musician who was featured on Kendrick Lamar's most recent album. I've been listening to Thundercat's new album on repeat while making some tasty plots in mathematica for my project. If you're curious, here's my favorite track by him:
Aaaaanyway,  plots! I had some spectacular tutelage in the art of plot-making back in the fall when Kieron, John, and I published that JCP article. I still use the mathematica code Aurora sent me and it always yields compliments. Being at the point in a project where I'm mashing data and words together in a TeX file feels pretty awesome. Few things are as satisfying as taking a step back and realizing how much stuff you've done over the course of a project.

Dan and I also got an opportunity to present some of our data during a little mid-summer report on Wednesday. The idea was that each group would spend 5 minutes discussing goals and means to achieve those goals before the summer was over. Since we already accomplished some stuff we'd aimed to do by August, we took the presentation as an opportunity to show off some of our results. In spite of being incredibly pressed for time, the talk went pretty well. I really wanted to go into details about the DFT we were doing because we're using the KSDT functional, also known as finite-temperature LDA! It's so cool, but we barely had enough time to talk about our results. I can't wait for the end-of-summer presentation where we'll have time to talk about some of the finer details of the calculations we've run.

Anyway, I should get back to plot-making. Listen to that song I posted. Seriously.