This is a great talk by @jeremyphoward about notebook-centered python development. Could be very useful in particular for writing scientific software. 1/n
I would like to emphasize, though, that this is mostly possible due to a tool Howard has written, nbdev, which turns notebooks into python packages. 2/n github.com/fastai/nbdev
I was always very skeptical about notebook-centered development, but I also didn't have nbdev, so I think my concerns were valid at the time. 3/n
I also think that many (but not all) of these issues are already addressed in the R community through the wide range of *down packages, such as bookdown, packagedown, blogdown, etc. 4/n
It's trivial to turn an R package into nice website, or write a book where all code is run and output is captured. 5/n
What we don't have in R is a way to convert a collection of .Rmd files into a package. That's an obvious opportunity somebody should go after. #rstats 6/n
The name {packagedown} is already taken, so maybe that package would have to be called {Rmddev}. 7/n
The end. 8/8
*Edit: It's {pkgdown}, not {packagedown}.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
This spring, I'm teaching a new class on data visualization with R. I'm posting all materials as I go. Feel free to follow along. Each lecture has slides and an interactive worksheet. #rstats wilkelab.org/SDS375/syllabu…
I'm using xaringan for slides, learnr for interactive worksheets, and distill for the website. All sources are available on github. github.com/wilkelab/SDS375
I'm about a third through the semester. New postings every week! 😄
First you subdivide all authors into two groups, people who have PI status (usually some sort of faculty member) and people who don't (grad students, postdocs, technicians, etc.). 2/n
Among the non-PIs, you sort them in order of their contribution, from most to least. 3/n
Spatial plotting just improved a lot in the development version of ggplot2. In a nutshell, you can now mix and match regular geoms with `geom_sf()` and `coord_sf()`. If you're doing any geospatial plotting, please test this out. 1/n #rstats#ggplot2
The key idea is that `coord_sf()` now has a default coordinate system that it uses for any objects that are not sf (i.e., don't come with their own coordinate information). The default is longitude/latitude, which makes it super easy to, for example, mark a specific location. 2/n
This works with essentially any geom, e.g., we can mark a couple of cities and draw an enclosing polygon. 3/n
"We have plenty of hospital bed capacity" is an awful argument for reopening in the face of a preventable infectious disease. For comparison, consider the following statements. 1/n houstonchronicle.com/politics/texas…
No need to wear a hardhat on this construction site. We have plenty of hospital bed capacity. 2/n
No need to wear a seatbelt. We have plenty of hospital bed capacity. 3/n