Zac Lawrence

[Zac Lawrence @zd1awrence]

@PGrasics @SimonLeeWx @wxiscool These plots will be fixed after today, but old ones will be wrong. I should also note that stratobserve is on its last legs, and I am working on a new and much improved successor, building on an additional 5 years of experience :)

[Zac Lawrence @zd1awrence]

@PGrasics @SimonLeeWx @wxiscool stratobserve plotting is done with *old* python 2.7 code/packages that haven't been touched for 5 years. Around 3 mos ago I changed the FFT package used for these calculations, but didn't read documentation closely enough on scipy.fftpack to understand its output format.

[Zac Lawrence @zd1awrence]

@shawnmilrad someone out there adding a shark

[Zac Lawrence @zd1awrence]

@DrAHButler @theAGU @NOAA_ESRL Congratulations, Amy! It's about time😉

[Zac Lawrence @zd1awrence]

@MarineShitposts @techsavvytravvy (sorry)

[Zac Lawrence @zd1awrence]

@SimonLeeWx @EarthSciStA @univofstandrews Congratulations, Simon! That's super exciting! Sounds like it will be a great fit

[Zac Lawrence @zd1awrence]

Then the regret sets in: you spent 15+ hours agonizing over this one review (with no compensation).

The "it's so over" -> "we're so back" -> "it's so over" cycle continues

[Zac Lawrence @zd1awrence]

TFW you're sure you're too dumb to peer review a paper and can't feel confident in giving a hard reject/accept. You submit your review with sweaty palms, certain that you're going to look like a fool to the editor, a distinguished expert.

[Zac Lawrence @zd1awrence]

After submitting, you're granted access to the other submitted review(s), and find that they agree with your own.

[Zac Lawrence @zd1awrence]

@MJVentrice @judah47 one thing likely going on here is a difference in methodology for computing the NAM/AO. it looks to me as if Judah's index standardizes across days of year, which will amplify small anomalies in late spring-summer. Compare attached fig to Fig 1c of https://t.co/Lr8mjPypDw

[Sol Osman @marisol_osman]

I'm happy to share my new paper with @cgrams_lsdp @dombueeler and @BeerliRemo on #S2S multi-model assessment of the year-round Atlantic-European regimes #LSDPatKIT https://t.co/7M0Z2eNz9D A brief 🧵:

[Zac Lawrence @zd1awrence]

@climatemodeler Better get in touch with CERN. We're detecting Higgs bosons in the upper stratosphere

[Zac Lawrence @zd1awrence]

@wxjay Forgot to mention that after you do the group by and mean/std on init.dayofyear, it will automatically get turned into dayofyear, giving you what you need

[Zac Lawrence @zd1awrence]

@wxjay So date[0] would be the init, and lead would be (date-date[0]). If you did this transformation to the data as you read in files, different forecasts could be appended on 'init' as a dimension. The rename and assign_coords methods for Datasets should do what you need, I think

[Zac Lawrence @zd1awrence]

@climatemodeler @SiyuanWang_PhD I'm curious to hear more specifics of the use-cases you've had success with. I've had some limited success with asking things like "write tests for these functions", but haven't yet had a need to try asking it to do something new (to me)