Showing posts with label science. Show all posts
Showing posts with label science. Show all posts

Sunday, April 26, 2020

Sunday: Q44/Descent into the Underworld.


I don't want to reorganize and caption these today.  The important thing that happened that made me stop playing the game is the first one:
Finding Phoibe.
I'll deal with that tomorrow.














I like Kassandra's snark.



I'm not tone deaf.




Wednesday, October 16, 2019

Wednesday: Tomorrow I will hopefully finish work projects.

Today had meetings and interruptions, and just wasn't good for work.  I also am in the "code integration" part of things, which is always hard.  It's like the eye-doctor's "A or B" tests, but with conflict markers.
This was easy.  The stack is deeper than the average of single visits, so there are more objects in that plot.  But, because of overlaps, you have to restrict the detections to only the primary detection in the overlap regions.

Tuesday, October 15, 2019

Tuesday: Sometimes I'm surprised when I remember how to science at things.

I think the pattern is mostly fake?

Because the image looks pretty good.  Although I don't really know if the scales line up completely.
And I can see the moon rise from my couch now.


Wednesday, August 14, 2019

Wednesday: Only half over?

That doesn't sound right.  I'm pretty sure I've been at this conference for forever at this point.  Still interesting, still learning stuff, still using things I've read about to make cool figures:
I can't figure out the transform code to flatten the orbits down into a plane, but I'm probably over looking something.
BBQ for the Team Dinner.  I like BBQ.

The gym.

And then some Codenames, which we lost the first game of because my team overthought.

Tuesday, August 13, 2019

Tuesday: "it’s not a special forces attack paras."

Second day of the conference, but the first with real sessions.  Plenaries, a session about the educational outreach (which I should probably sent suggestions and code pointers for), one on how we're going to handle calibration products (sanely), and one about calibration (which had no chairs).

Then there was the "banquet", which I ditched out of early because it was hot, loud, and full of people.

I saw a lizard as I was leaving.
I then took a quick nap, and then looked into how hard it would be to code up some simple cuts and cleaning for PS1DR2 data.
Turns out, pretty easy.
I didn't implement a background source removal, but that's not super-too-hard to do (generate likely contamination probabilities by comparing source counts per sky area per plot area, and then selectively show sources depending on that probability).  Also, white dwarfs are more orders of magnitude fainter than I remembered.



Thursday, August 1, 2019

Thursday: This is why documentation is super important.

Because with a full set, it wouldn't have taken me so long to make this:
So if I have time tomorrow, I'll look into adding the overlay stuff to the widget code I was working with.  And then there'll be a telecon before we head home.  And then home.

And then serious "my own bed" ptfo.

Friday, July 12, 2019

Monday, July 8, 2019

Monday: Conference tutorials.

I have learned.  In particular, that we are doing many things well at work.


Tuesday, June 18, 2019

Tuesday: "Wait! I have a gun in this game!"

It certainly makes it a lot easier to beat the giant woman boss of level one if I can shoot her instead of relying only on my slowly regenerating magic.  In any case, Bloodstained is about as good of a game as I was hoping it'd be, and even though it suffers a bit from late-IGA "lots of bits and collecting and crafting and decision trees", it's still excellent.  Having Yamane Michiru and Kojima Ayami doing music and art helps give it the proper Castlevania attitude, even if they had to file off all the serial numbers since Konami is dumb.

It lets you dress up your character with silly hats and masks.  The NPCs naturally dress themselves in Church Appropriate Attire.

You can sit in chairs.
So, good game.

I also sorted out the first big block on my new work project, got some additional viewpoints on it, and then blocked out my next month or two.
It's probably not that bad, the major issue is getting buy in for all the changes on the orange post-it.


Friday, May 31, 2019

Friday: I finally turned on my A/C.

It was just too hot this evening.  I could probably go back to open windows tomorrow.

Today was a pizza-for-dinner day.
And I solved a bunch of major issues with the new robospect code, so it now actually fits things when I tell it to.  There's some double-counting issue to sort out, but the majority of functions should be working now.

I just have to merge it all, and then open tickets for the infinity things that need to be done to make it "done."  Still, it's good to know that the main problem to finishing up some of this has been "typos".



Wednesday, May 29, 2019

Wednesday: I never seem to get as much done as I would like to get done.

At least today I helped close out two things, and I think I've discovered the secret hidden bug in another.  I guess if I can get it all done by the weekend, that'll be worthwhile.

How????

We had these clouds today.
And then it rained a whole bunch, and I didn't have an umbrella, and I got wet.  :(



Thursday, January 31, 2019

Thursday: An ok day.

Tomorrow is demo day, and I got all my code in in time.  I also talked to the doctor people, and they should be doing my prescription, so hopefully that'll be sorted out by next week.

And this mostly is on the way towards working.  It's just coding up algorithms and making it "user friendly" at this point.

Monday, January 28, 2019

Monday: I'm annoyed that we don't already have a function to deal with this issue.

At least the second data release was finally published.

Sunday, January 20, 2019

Sunday: The riddle of queso

It's challenging, because it is easy to make, and yet you probably shouldn't eat a giant bowl of queso in one sitting.  Of course, now I'm hungry again, and keep thinking about it, sitting in my fridge, waiting for me to eat it.

And that's pretty much how my today went.  I woke up, I watched a bunch of youtube stuff, pushing my watch later list under 250, made queso, ate queso, watched Supergirl, played video games.  It was an excellent day, although I probably should have done laundry, and I probably should have put together the bookshelf I got yesterday.

My phone can't really capture the dynamic range needed for an eclipse, but there was an eclipse tonight.

Tuesday, April 24, 2018

Tuesday: A work at home day.

Mostly because it's "cold" and kind of rainy, and that means when I woke up this morning, all my joints decided not working was a fun prank to pull.

But I was able to prove that I almost have this thing working.  The yellow bars are supposed to be over the blue bars.  It doesn't look like a simple scaling, so I'll have to try something different tomorrow.

Tuesday, February 27, 2018

Tuesday: I think this is close to a solution.

I was expecting a point, but whatever.

  • That's depressing.
  • Legends of Tomorrow is still the best CW/DC show.  Tonight the Flash screwed up a bunch of times and got somebody killed.  Black Lightning beat up his daughter because she was dressed up like a super hero, and a smudge of makeup was confusing.  LoT went back in time and fought pirates.  I mean, this isn't even a battle anymore.

Monday, February 26, 2018

Monday: Ugh.

I think the talk is done.  Or at least I'm done with it.
And then I have to solve this problem tomorrow.

Friday, February 23, 2018

Friday: Conveniently, everything worked out.

Two projects mostly wrapped up today.  That's always a good thing.
To be fair, it's kind of hard to write any program with CFITSIO that isn't a dumpster fire.  I eventually decided it wasn't worth my time to figure out why it couldn't read strings correctly, and just manually added the strings myself.
But the "plausible output file" was actually a totally fine result, and it worked without much kicking.

So this jerk is my first shiny Pokemon.