Elektrine
Log in Register
Paige Chat Timeline Gallery Friends Email Drive DNS Private DNS Domains VPN Kairo Nerve
Remote

Erik-Jan

@erikjan@fosstodon.org
mastodon 4.7.3
  • Open on fosstodon.org

I work at bol (bol.com) on causal inference, statistics, measurement, recommendations, and machine learning for online advertising ranking and pricing.

Previously, I was assistant professor at Utrecht University, where I founded the ODISSEI Social Data Science team and worked on a broad range of topics in social science, statistics, and computation.

#statistics 📊, #rstats :rstats:, #julialang :julia:, #python :python:, #stan

Posts in Dutch and English

0 Followers
0 Following
9 Posts
Joined December 24, 2022
website:
https://erikjanvankesteren.nl
GitHub:
https://github.com/vankesteren
Open post
Erik-Jan @erikjan@fosstodon.org
· 4mo ago

So I moved into industry 1.5 months ago, which has meant a proper switch from R :rstats: to Python :python: (I love both). Here are a few observations for statistics-related stuff in this switch (mainly GLMs, statistical inference, contrasts)

- #polars is really great, I love LazyFrames & streaming millions of rows of parquet files, categorical data, missing data.
- I don't really like using #statsmodels, the interface is clunky and the formula API is unfinished

1/n

#DataScience #Statistics

fosstodon.org
5
1
2
0
Open post
Erik-Jan @erikjan@fosstodon.org
· 4mo ago

https://kanikeenkortebroekaan.nl

kanikeenkortebroekaan.nl
2
0
4
0
Open post
Erik-Jan @erikjan@fosstodon.org
· 5mo ago

My son got a second-hand R/C car today on #kingsday and of course it did not work after we put the batteries in. But secretly I like that a lot because I get to look inside and try to repair it 👨‍🔧

fosstodon.org
1
0
0
0
Open post
Erik-Jan @erikjan@fosstodon.org
· 5mo ago

RE: @erikjan@fosstodon.org

oh, saw now that that I should tag this with #NoAI

fosstodon.org
1
0
0
0
Open post
Erik-Jan @erikjan@fosstodon.org
· 6mo ago
Replying to
@emjonaitis procedures like the dbreg package (https://grantmcdermott.com/dbreg/) basically do the following: summarize the rows into buckets of profiles, and then perform weighted regression (weighted by how many people fall in each profile) This might be possible for you? Even though there are many rows there might be an order of magnitude fewer unique rows. You would need to bin your continuous predictors, so this is also approximate And not sure if this will even work with the nested structure
grantmcdermott.com

dbreg

1
0
0
0
Open post
Erik-Jan @erikjan@fosstodon.org
· 6mo ago
Replying to
@emjonaitis oh yeah, you're right, forgot that this is not in betareg
1
0
0
0
Open post
Erik-Jan @erikjan@fosstodon.org
· 6mo ago
Replying to
@emjonaitis this is a stretch, but do you think BRMS / Stan with Laplace approximation will be any faster?
1
2
0
0
Open post
Erik-Jan @erikjan@fosstodon.org
· 44mo ago
Replying to
@ropensci @jouni_helske @lucystats @rstats this looks great!
2
0
1
0
Back
313k7r1n3
Elektrine

Tor hidden service

elekhj7afj4qnrr4yd3bkzslsyo5jgfxw3orgjkhlcxifueodybyiiad.onion

I2P eepsite

j6b6cyk6gjmepjih7jjadxgxvvf3lzzujljuu2v4biemzpg3naya.b32.i2p

Platform

  • Email
  • Chat
  • Timeline
  • VPN
  • DNS

Company

  • About
  • Contact
  • FAQ
  • Lite (no JS)

Legal

  • Terms of Service
  • Privacy Policy
  • Transparency Report
  • Report Abuse
  • Warrant Canary
  • VPN Policy

Support

  • support@elektrine.com
  • Report Security Issue
Mail client setup IMAP mail.elektrine.com:993 POP3 mail.elektrine.com:995 SMTP mail.elektrine.com:465
© 2026 Elektrine. All rights reserved. Server: 22:30:24 UTC