Sign in

Steve Ewing

@jambuddy.live
1.7K followers 2.9K following 593 posts

Data munging dad in GA. Playing with music, code or both at the same

PostsRepliesMedia
Steve Ewing @jambuddy.live · 28/03/2026
Liberty and Justice for Y'all #NoKings Decatur, GA
The statue of john lewis in decatur square with a kids protest sign reading Liberty and Justice for Y'all
051
Steve Ewing @jambuddy.live · 13/03/2026
The fools should have remembered Vizzini
Vizzini from Princess Bride and the quote, "Fool! You're falling victim to one of the classic blunders... Never get involved in a land war in Asia!
020
Steve Ewing @jambuddy.live · 03/02/2026
Alright there algorithm, you've been cool about showing me data science and book posts but I need more dogs and luthiers in my feed!
Old dog in workshop with the shell of a mandolin
010
Steve Ewing @jambuddy.live · 15/01/2026
Here's a project I can get behind
000
Steve Ewing @jambuddy.live · 22/04/2025
This is why no1 uses you Google. Can you even imagine Microsoft doing this?
040
Steve Ewing @jambuddy.live · 05/12/2024
That's a good one, learned something! Thanks for sharing
A chart combining box plots with error bars and violin plots by splitting them down the middle and setting them next to each other
010
Steve Ewing @jambuddy.live · 21/11/2024
When you use it for row grouping is there a way to have a new column with the group outline and a label using tt?
100
Steve Ewing @jambuddy.live · 19/11/2024
For example, if you make a map using the Zip Code Tabulation Areas from tidycensus for the whole US it'll be too big to upload to github and they'd need to figure out how to soft reset their commit
020
Steve Ewing @jambuddy.live · 18/11/2024
This is a fun dataset! It'd be good for teaching this. #Rstats
The image displays the results of a GLM model predicting rating with dialogue_density, avg_length, and exclamation_ratio as predictors.

The intercept is 
7.69464
7.69464, significant (
𝑝
<
2
𝑒
−
16
p<2e−16).
dialogue_density has a positive coefficient of 
0.15753
0.15753, significant (
𝑝
=
0.000661
p=0.000661).
avg_length has a negative coefficient of 
−
0.16881
−0.16881, significant (
𝑝
=
0.000461
p=0.000461).
exclamation_ratio has a positive coefficient of 
0.14059
0.14059, significant (
𝑝
=
1.97
𝑒
−
06
p=1.97e−06).
Residual deviance: 
46.712
46.712 (on 
267
267 degrees of freedom).
Null deviance: 
53.244
53.244 (on 
270
270 degrees of freedom).
AIC: 
302.61
302.61.
Overall, the model highlights three significant predictors, with the residual deviance indicating the unexplained variance.
262
Steve Ewing @jambuddy.live · 18/11/2024
Twilight
r studio appearance settings
000
Steve Ewing @jambuddy.live · 18/11/2024
Very cool. Any idea how one can make it also work in dark mode?
his same map in dark mode showing white text on a white background
100
Steve Ewing @jambuddy.live · 17/11/2024
Reposted by the father of cyberpunk himself.
the same as op just with William Gibson's repost
110
Steve Ewing @jambuddy.live · 16/11/2024
Ha, it worked. 🍻
000
Steve Ewing @jambuddy.live · 16/11/2024
Haha yeah right, I wish I was this cool
graphic from skircle.me that shows a lot of data science rock stars and politicos.
010
Steve Ewing @jambuddy.live · 16/11/2024
It knows... "Based on what you know about me make an image about what you think my life looks like."
110
Steve Ewing @jambuddy.live · 01/11/2024
Last nights festivities have left this dragon feeling slain this afternoon
Inflatable dragon costume and twin 4yo knights
010
Steve Ewing @jambuddy.live · 31/10/2024
Love the ending message of this talk. Thanks for sharing it Malcom!
030
Steve Ewing @jambuddy.live · 24/10/2024
Watched this on Rami's recommendation. It was amazing to see the state of the art in time series forecasting. The way he forecasts sparce series by splitting it into two forecasts is in hindsight so obvious... I'm also going to try my hand at Hierarchical Pricing Elasticity it's a brilliant idea.
051
Steve Ewing @jambuddy.live · 21/10/2024
I think it would help to give an example. The last 4 lines of the first paragraph. They are using the demographics to choose their sample. But when its a divide along the same demographic lines they are segmenting on, like it was in 2016, doing so puts the cart before the horse.
000