Andy Berner
@andybrnr.bsky.social
Meteorology PhD, data scientist, ski mountaineer, GSP dad
Wonder how this holds up - never read it, just vague recollections of the movie.
September 14, 2025 at 5:36 AM
Wonder how this holds up - never read it, just vague recollections of the movie.
Literally this meme they love to post, but reality is 180 degrees out of phase with what they think it is.
August 26, 2025 at 7:58 PM
Literally this meme they love to post, but reality is 180 degrees out of phase with what they think it is.
Useful figure (from previous link) comparing protest event count (not attendance) vs 2017:
August 16, 2025 at 8:44 PM
Useful figure (from previous link) comparing protest event count (not attendance) vs 2017:
It really does. Same TH, blueberries in mid-August, pow turns in February.
August 16, 2025 at 12:07 AM
It really does. Same TH, blueberries in mid-August, pow turns in February.
Not exactly standard Western WA Wx for mid-August, but I’ll take it.
August 15, 2025 at 10:57 PM
Not exactly standard Western WA Wx for mid-August, but I’ll take it.
7.5yo GSP and 14mo Eurohound are best buds :)
August 7, 2025 at 8:34 PM
7.5yo GSP and 14mo Eurohound are best buds :)
C-17 HUSKY25 on approach to Whidbey NAS.
July 30, 2025 at 3:54 AM
C-17 HUSKY25 on approach to Whidbey NAS.
More SkiMo than climbing the last few years, but trying to get back into it!
July 21, 2025 at 5:44 PM
More SkiMo than climbing the last few years, but trying to get back into it!
Think that marks the circulation center for a tropical disturbance (from www.vos.noaa.gov/MWL/april_05...)
July 5, 2025 at 4:29 PM
Think that marks the circulation center for a tropical disturbance (from www.vos.noaa.gov/MWL/april_05...)
Yeah, 100% that's plotting the logit model output. The original data is just 25 cols of metadata coding that codes an event as success, limited or failure. If I just use the success column and log scale the pct of pop participating (actually the frac of pop, but meh), I get this:
June 15, 2025 at 6:27 PM
Yeah, 100% that's plotting the logit model output. The original data is just 25 cols of metadata coding that codes an event as success, limited or failure. If I just use the success column and log scale the pct of pop participating (actually the frac of pop, but meh), I get this:
Not trying to nitpick, but wouldn't "housing cost" encompass both financing cost and price level? COVID flows did some pretty wild things to regional market prices...
April 1, 2025 at 5:55 PM
Not trying to nitpick, but wouldn't "housing cost" encompass both financing cost and price level? COVID flows did some pretty wild things to regional market prices...
