Simon Jackson

@drsimonj

Lead Data Scientist at , empowering people to make good decisions. Love to talk stats, psych, decision making, machine learning, R

Amsterdam, The Netherlands
linkedin.com/in/drsimonj
2015(e)ko martxoa(e)tik Twitterren
81 Argazki edo bideo Argazkiak eta bideoak

Txioak

@drsimonj blokeatu duzu

Ziur zaude txio hauek ikusi nahi dituzula? Txioak ikusteak ez du @drsimonj desblokeatuko.

  1. Finkatutako Txioa
    abu. 21

    "Exploring correlations in R with corrr" A blog-post version of my amst-R-dam meetup talk. BONUS: Features DB backend addition by

    2 erantzun 34 bertxio 120 atsegin
    Desegin
  2. (e)k Bertxiotua
    abe. 8

    🎁 Day 8 👨‍⚕️ Simon Jackson () 💪 {corrr} for tidy correlations 🔧 {twidlr} for tidy model pipelines 📝 Great blog posts

    Exploring correlations in R with corrr
    2 erantzun 66 bertxio 232 atsegin
    Desegin
  3. (e)k Bertxiotua
    aza. 28

    At long last, the parsnip package is on CRAN. parsnip is a unified R interface to models. The first of two blog posts about the package is at

    15 erantzun 277 bertxio 754 atsegin
    Desegin
  4. (e)k Bertxiotua
    aza. 27

    🎞 these are so cool! "causalgraphs: R code to generate animated displays of basic causal inference methods" 👨‍💻

    2 erantzun 49 bertxio 177 atsegin
    Desegin
  5. (e)k Bertxiotua
    aza. 15

    Materials from my with and talk at are now available here:

    2 erantzun 29 bertxio 80 atsegin
    Desegin
  6. (e)k Bertxiotua
    aza. 14

    🤩 TIL about reticulate::py_install(), meaning I can forget about all these various ways to install 🐍 stuff.

    Erantzun 1 7 bertxio 38 atsegin
    Desegin
  7. (e)k Bertxiotua
    aza. 11

    “Any fool can write code that a computer can understand. Good programmers write code that humans can understand.” – Martin Fowler

    12 erantzun 665 bertxio 2.019 atsegin
    Desegin
  8. (e)k Bertxiotua
    aza. 9

    Hey , want to brush up on your tidyeval? Try this interactive tutorial covering symbols, quosures, !!, !!!, and more: (feedback welcome)

    17 erantzun 175 bertxio 686 atsegin
    Erakutsi haria
    Erakutsi haria
    Desegin
  9. (e)k Bertxiotua
    aza. 8

    Great advice from : if you want to get started with , use a dataset you’re interested in, not mtcare. One possibility - download your social media or gmail data!

    2 erantzun 16 bertxio 59 atsegin
    Desegin
  10. (e)k Bertxiotua
    aza. 5

    I'm starting monthly bite-size data posts on meta! 📊 Check out this first one about how users' actions are correlated.

    3 erantzun 17 bertxio 58 atsegin
    Erakutsi haria
    Erakutsi haria
    Desegin
  11. (e)k Bertxiotua
    aza. 4

    😻 Just saw 's work for the first time, and I'm so in love! 🖍 "Sketchplanations - A weekly explanation in a sketch"

    Mercator Projection
    3 erantzun 50 bertxio 178 atsegin
    Desegin
  12. (e)k Bertxiotua
    urr. 30

    Find a typo somewhere in a bookdown book (e.g. ) or in a package readme/docs on github? You can propose the fix yourself! Follow this step-by-step tutorial w/ lots of screenshots . H/t & thank you to .

    31 bertxio 83 atsegin
    Desegin
  13. (e)k Bertxiotua
    urr. 17

    hey! I just gave a talk at on what I've discovered about while hosting , the podcast. Slides are here:

    5 erantzun 27 bertxio 117 atsegin
    Desegin
  14. (e)k Bertxiotua
    urr. 17

    Great talk from yesterday on the benefits of coding your data analysis for reproducibility. Bonus: I’ve never related to a speaker more. “All code is collaborative even if it’s just between you and future you”

    4 erantzun 63 bertxio 334 atsegin
    Desegin
  15. (e)k Bertxiotua
    urr. 17

    Just like reserving money for unforeseen costs when budgeting, you should set aside time for unforeseen delays when planning for a data science task. Reduces stress when something keeps failing without knowing why or you just don't see the answer right away.

    5 erantzun 9 bertxio 31 atsegin
    Desegin
  16. (e)k Bertxiotua
    urr. 12

    Materials from my with are now available in this repo: , Includes code, deck and links to the published work.

    15 bertxio 56 atsegin
    Desegin
  17. urr. 10

    "All the right math and data to solve the wrong problem" - sharing that type III errors are at the heart of Decision Intelligence

    11 bertxio 38 atsegin
    Desegin
  18. urr. 10

    "Irrationality is an expensive bug" - I love this setup of human decision biases and how / can improve decision making by at

    eta
    Erantzun 1 7 bertxio 5 atsegin
    Desegin
  19. (e)k Bertxiotua
    urr. 8

    Just tinkering… 📊 "Pretty scatter plots w/ ggplot2" 📽 "Decouple Code and Output in xaringan"

    2 erantzun 40 bertxio 131 atsegin
    Desegin
  20. (e)k Bertxiotua
    urr. 5

    Running the tests again without saving your changes

    147 erantzun 8.132 bertxio 15.867 atsegin
    Erakutsi haria
    Erakutsi haria
    Desegin
  21. urr. 7

    ICYMI. How and when: ridge regression with glmnet Or, "When does Ridge🥊 OLS?" 🔗

    Erantzun 1 7 bertxio 56 atsegin
    Desegin

@drsimonj(e)k ez du oraindik Txiokatu

Badirudi kargak luze hartuko duela.

Baliteke Twitterren kapazitatez gaindi egotea edo momentuko gainkarga bat izatea. Berriro saiatu edo Twitterren Egoera ikusi informazio gehiagorako.

    Agian hau ere gustuko duzu

    ·