David Robinson

@drob

Chief Data Scientist at , fan/evangelist

New York, NY
varianceexplained.org
2009年6月加入
1,377 個相片或影片 圖片與影片

推文

你已封鎖 @drob

你確定要查看這些推文嗎？查看推文並不會將 @drob 解除封鎖

  1. 已轉推
    11 小時前

    I'll be speaking at about a streamlined workflow for tidy time series analysis using tsibble and fable. A warm-up post to provide an overview of tsibble with a case study of NYC Citi bike data.

    31 則轉推 106 個喜歡
    還原
  2. 24 小時前

    If Trump actually did shoot someone on Fifth Avenue, the headline would be “In midtown violence, echoes of a campaign boast”

    7 則轉推 65 個喜歡
    顯示此對話串
    顯示此對話串
    還原
  3. 24 小時前

    NYT reporters are under a Cassandra-like curse where they get amazing scoops but have to hide it behind vaguely dramatic headlines

    This is not the main point but it is amazing that the NYT hed for "Facebook let Netflix and Spotify spy on your DMs" is "As Facebook Raised a Privacy Wall, It Carved an Opening for Tech Giants"
    顯示此對話串
    1 replies 6 則轉推 14 個喜歡
    顯示此對話串
    顯示此對話串
    還原
  4. 已轉推
    12月17日

    Happy to announce that we have some internship positions available in statistics and machine learning at Netflix for next summer. Visit this posting to learn more:

    1 replies 63 則轉推 84 個喜歡
    還原
  5. 12月17日

    Nothing beside remains. Round the decay Of that colossal Wreck, boundless and bare The lone and level sands stretch far away.

    3 replies 6 則轉推 63 個喜歡
    還原
  6. 已轉推
    12月17日

    🦕 My latest weekend project is a game to guess which tweets are from Kanye and which are from Elon. ➡️ 😜 It’s harder than it sounds. 🎉 Good luck.

    29 replies 231 則轉推 670 個喜歡
    還原
  7. 已轉推
    12月15日

    “There’s already 20 million people who can build software - isn’t that enough?” No. Until making your own software is as accessible as writing your own book, we’re not there yet. Not everyone can (or wants to) write a great book, but most have the baseline skills to try.

    14 replies 48 則轉推 316 個喜歡
    還原
  8. 12月15日

    I have such a common and easy name how did I fuck this up

    10 replies 16 個喜歡
    顯示此對話串
    顯示此對話串
    還原
  9. 12月15日

    The barista heard my name as Gabe and I didn’t correct her. I’m going back for seconds and heaven help me I’m going to have to say Gabe again – 在 Joe Coffee Company

    6 replies 1 則轉推 23 個喜歡
    顯示此對話串
    顯示此對話串
    還原
  10. 已轉推
    12月14日

    About to get a haircut, time to make a vague statement about what I want and immediately agree to however they interpret it.

    52 replies 1,136 則轉推 9,144 個喜歡
    還原
  11. 已轉推
    12月14日

    it's called a confusion matrix because you can never remember what the rows and columns mean

    9 replies 270 則轉推 1,466 個喜歡
    顯示此對話串
    顯示此對話串
    還原
  12. 12月14日

    "You'll have plenty of time to get kidnapped after your manuscript is submitted"

    A Swedish-Iraqi student was rescued from an ISIS held town sometime in 2014 by mercenaries in four SUV’s hired by his thesis advisor and we’re only just hearing about it. 😬
    顯示此對話串
    1 replies 16 則轉推 81 個喜歡
    還原
  13. 已轉推
    12月14日

    "Why my code crash?" "Sir, I'll need a self-contained reproducible exa--" "WHY MY CODE NO RUN?!!"

    4:11
    Just remember, these are elected officials and our tax dollars are being used for this
    6 replies 50 則轉推 247 個喜歡
    還原
  14. 已轉推
    12月13日

    monty hall problem solved in . Inspired by 's screen casts

    1 replies 10 則轉推 53 個喜歡
    還原
  15. 12月13日

    In the orientation for my PhD program, a professor recommended a biology class, but added “It’s probably too hard for someone whose background is, say, statistics.” I was the only stats major in the program, which he knew. He did not like me

    Tweet your favorite diss while in grad school. Mine: I went to my mathematical logic professor for help understanding the proof of a complicated theorem. He replied: “Don’t feel bad, Scott. Even talented students have problems with this.”
    1 replies 6 則轉推 65 個喜歡
    還原
  16. 已轉推
    12月11日

    I've been using travis for quite some years, and this is still what my commit history looks like when trying to set up a new configuration

    7 replies 20 則轉推 180 個喜歡
    還原
  17. 已轉推
    12月11日

    An open letter to , , and regarding algorithms and my son's birth:

    2,479 replies 28,388 則轉推 67,318 個喜歡
    顯示此對話串
    顯示此對話串
    還原
  18. 12月12日

    Here's an alternative solution to the problem that's much less readable but fits in a tweet: mean(replicate(1e5, all(between(cumsum(rbinom(10, 1, rep(c(.5, .4), 5)) * rep(c(-1, 1), 5)), -2, 1)))) only non-base function is dplyr's between()

    In this 20 minute bonus screencast, I answer a probability puzzle from last Friday's "Riddler" column using Monte Carlo simulation in 🎲 It's hard to solve puzzles under pressure 😅
    2 replies 5 則轉推 22 個喜歡
    還原
  19. 12月12日

    In this 20 minute bonus screencast, I answer a probability puzzle from last Friday's "Riddler" column using Monte Carlo simulation in 🎲 It's hard to solve puzzles under pressure 😅

    1 replies 18 則轉推 130 個喜歡
    還原
  20. 已轉推
    12月12日

    📕 If you or a friend have a kid learning to 📖 read 📖 get a deck! (we can still ship in time for xmas) 📕

    2 replies 6 則轉推 4 個喜歡
    顯示此對話串
    顯示此對話串
    還原

@drob 尚未發表任何推文。

看來要一段時間讀取資料。

Twitter 可能已超出負載，或發生暫時性的小問題。請再試一次，或造訪 Twitter 狀態以取得更多資訊。

    你也可能也會喜歡

    ·