ijlyttle

SSH for travis-pkgdown

In the .ssh directory, we should have:

id_rsa & id_rsa.pub - the default keys for this computer

In addition, you need a distinct set of keys for each repository. Accordingly, each of these would be a directory within .ssh:

• gh-deploy-vegawidget
• gh-deploy-altair

ijlyttle

Motivation

I see our project as a set of two transformations: one that changes a ggplot2 object into a ggspec, then another that changes a ggspec into a Vega-Lite spec.

The analogy may not be exact, but I see these transformations in terms of linear algebra, where a ggplot2 object is a vector in "ggplot" space, a ggspec is a vector in "ggspec" space, and a Vega-Lite spec is a vector in "Vega-Lite" space.

One of our goals is that "ggspec"-space should be a faithful representation of "ggplot2"-space. One way of doing this is to make sure that the "transformation-matrix" is as close to diagonal as we can make it. As Haley likes to point out, the ggplot2 object is a "list of 9"; therefore the "ggspec" object will have no more than 9 elements (maybe it will not have a "theme"?).

If we wanted to (and we don't want to), we could reproduce the ggplot2 object using the ggspec.

ijlyttle

# An .Renviron file helps keep you safe, by storing secrets in one place that 
# you can secure.
# - to open *your* .Renviron file, usethis::edit_r_environ()
# - please make sure that your .Renviron file ends with a newline, 
#   or R will ignore it
# - this is a minimal file, you will likely want to add your own "stuff"
#   - if you need to deal with a firewall, this is the place to define 
#     your proxy variables
#
# More information: https://whattheyforgot.org/r-startup.html#renviron

ijlyttle

gistr

ijlyttle

license: mit
height: 500
scrolling: yes
border: yes

ijlyttle

license: mit
height: 500
scrolling: yes
border: yes

ijlyttle

license: mit
height: 500
scrolling: yes
border: yes

ijlyttle

license: mit
height: 500
scrolling: yes
border: yes

ijlyttle

library("rlang")
library("dplyr")
library("purrr")

myfil <- function(data, names, vals) {
  
  exprs <- purrr::map2(names, vals, ~ quo(!!sym(.x) == !!.y))  
  
  print(quo(dplyr::filter(data, !!!exprs)))
  

ijlyttle

Include something like this in your Rmd to "turn-off" the borders for all accordion displayed content-panels.

```{css}
.panel-success {
    border-width: 0;
}
```