non-global binding and R check

[mwpennell]

I am using some dplyr in some functions and am using variables within a function in a few places. e.g.:

add_binomial <- function(df)
    df %>% rowwise() %>%
        mutate(resolved_binomial = short_species_name(resolved_name))

R CMD check throws a note:

add_binomial: no visible binding for global variable ‘resolved_name’

is this going to be a problem for CRAN? Is this good coding practice. This seems like a recurring issue in the hadleyverse

[cboettig]

tl;dr I think you want mutate_ not mutate

Yeah, many functions Hadley writes use non-standard evaluation as a way of giving the user a slightly more concise input (e.g. not quoting variable names). This results in syntax that easily confuses R (e.g. is resolved_name a string or a variable containing a string?) and Hadley recommends against using this syntax when writing packages. All of the dplyr functions include an equivalent version of the function using standard evaluation syntax. The standard evaluation version is just the original version with an underscore at the end. See http://cran.r-project.org/web/packages/dplyr/vignettes/nse.html for a proper explanation.

• Carl

p.s. this stuff still confuses me too.

On Tue Dec 16 2014 at 12:34:58 PM Matt Pennell notifications@github.com wrote:

I am using some dplyr in some functions and am using variables within a function in a few places. e.g.:

add_binomial <- function(df) df %>% rowwise() %>% mutate(resolved_binomial = short_species_name(resolved_name))

R CMD check throws a note:

add_binomial: no visible binding for global variable ‘resolved_name’

is this going to be a problem for CRAN? Is this good coding practice. This seems like a recurring issue in the hadleyverse

— Reply to this email directly or view it on GitHub https://github.com/ropensci/chromer/issues/16.

[mwpennell]

Alright. Thanks -- this is weird, but it is now fixed (I think).