seankross
quality-of-life: display Vectors, model objects, and other 'unsupported' objects as strings (but beware of length)
not sure if this ever comes up in %>% pipelines, but sometimes in pandas the result of a pipeline is a 1-D Series object (as opposed to a 2-D DataFrame), so if that comes up too in R, we can support those as well. but again, no need to sweat it.
Re - scalars. Sean: There are no scalars in R, only vectors of length 1. So whatever we do to handle vectors will also cover "scalars"
rhs is a vector here, i believe? right now we just punt and display nothing on rhs (actually right now we crash but after i fix things it will display nothing on rhs)
library(dplyr)
library(palmerpenguins)
penguins %>% pull(species)
if we don't want to get into this complexity yet, a thing that pandas_tutor does is simply print out the string form of non-dataframe-y things. so just print the vector as a string and call it a day. that way, it simulates what the R console prints out. (however, we have to be careful about LONG vectors that take up a ton of space, and account for that in your 1MB upper bound cap ... or else if someone creates a huge vector then it will crash the server)
The same thing happens if you pipe into lm() or first() ... since we don't want to support arbitrary objects like linear model objects, etc., i think the most graceful thing is to print them as a string summary, like whatever would've been printed to the R console