I missed hanging out with @CVWickham who I love chatting with and scheming about nerdy acronyms and marine puns with @gavin_fay too. And congratulations - @hadleywickham is a very happy and proud uncle 💕
What an amazing few days. I have so many wonderful messages I can't wait to catch up on and moments to share. Here is a start: today I'm w @eeholm at @NOAA HQ and we've been co-working in @quarto_pub 🌈
@healthandstats Feeling the same. I now realise I should have blocked out the entire day in my calendar, like I would have if I had to travel to be there.
@CVWickham I also like that one, although less "metaphoric". I can't remember exactly, but I think its from a presentation/blog-post by @JennyBryan or @CVWickham...
Do you have a favourite metaphor for subsetting lists in #rstats? I'm collecting links for my students. So far, I've got the condiments metaphor (#lists-of-condiments" target="_blank" rel="nofollow noopener">r4ds.had.co.nz/vectors.html#l…) and the train metaphor (#subset-single" target="_blank" rel="nofollow noopener">adv-r.hadley.nz/subsetting.htm…).
Doing your part to stay away from everyone? Treat yourself by watching @CVWickham first @pluralsight course "Building Your First R Analytics Solution" bit.ly/2QKXpeG. Happy to set up anyone with a 30-day free trial. #rstats
I am not sure how to interpret this: "Fit regression model, use differenced series, or specify ARIMA(0, d, 0) errors." Fit regression model to differenced series x_t and y_t assuming ARIMA(0,0,0) errors? Fit regression model to original x_t and y_t assuming ARIMA(0, d, 0) errors?
@CVWickham: Charlotte, I am reading your excellent slides on spurious regression (stat565.cwick.co.nz/lectures/12-sp…) and am not sure I am interpreting your "Strategy for Regression" correctly.
@IsabellaGhement I'm hesitant provide a concrete answer to this, since my head has been out of this material for too long. My intuition says yes, but I'd check by simulating and testing it out.
If y_t and x_t are non-stationary (e.g., both exhibit a linear trend), can we fit a model of the form y_t = beta0 + beta1*xt + e_t to the original time series which allows e_t to be ARIMA(0,d,0) and examine the residuals r_t of this model for ARMA(p,q) structure?
Upcoming webinar: Teaching Online on Short Notice
@gvwilson shares advice for moving your in-person class online in a hurry.
March 19th at 4PM EDT
Learn more and sign up at, resources.rstudio.com/upcoming-webin…#rstats
Are you an #rstats user in the PNW? Want to share your work with fellow researchers/data scientists? Submit a proposal for a 5-min lightning talk for the 4th-annual #CascadiaRConf in Eugene, OR! Call for lightning talks is open through April 15th, 2020: bit.ly/2Tdskke
Happy to announce that Lightning talk submissions for the 2020 #CascadiaRConf are now open! Come visit us in Eugene and share the ways you use #rstats! Applications open through mid April. docs.google.com/forms/d/e/1FAI…
RStudio's 2020 summer intern program is now accepting applications: please see buff.ly/2T8sFVa for the announcement and buff.ly/2wJHoyn for the application form.
#rstats