Nomenclature for variables/columns/attributes

[hermands]

@jwrudolf @amromeo Sorry if I missed this earlier.

• 01 is referring to columns as 'attributes', but rstudio writes as 'variables'
• 03 is using 'columns'

I was going to use 'variables'... What is our convention?

[pcmathias]

I think we should follow the R for Data Science terminology and refer to columns as variables and rows as observations.

[amromeo]

My personal preference would be column. I think thats what our audience would call them. R/tidyverse isn't really consistent with these things. select() does use variables, but what about cbind() bind_cols()

R4ds says something like: in tidy data Each variable must have its own column, each observation its own row so to me variable and observation are ideas in the data, a column is that vertical thingee in a data frame.

anyone convinced?

[pcmathias]

I think it's fine to use both and agree that column is the physical description of the data structure and variable is an idea. I think we just want to avoid multiple terms for the idea terms.

[amromeo]

that works for me as well

[jwrudolf]

Sounds good.

I changed my language to variables & columns, rows & observations.