paper/jenny_test_drive.md

In Ironholds/practice: Tools for identifying whether R packages follow common software engineering conventions

Jenny Test Drive

Jenny Bryan r Sys.Date()

Temporary workaround: date info

Until CRANpractices gets rebuilt with more info about a package's birthdate, most recent update, and number of releases ... I'm gonna grab what I can from CRANmetadata.

## Warning in left_join_impl(x, y, by$x, by$y): joining factor and character
## vector, coercing into character vector

Frequency tables and barcharts

Many variables take on a small number of values and we can digest their frequencies en masse.

There are 6551 packages in this dataset.

Variables in CRANpractices that are not covered above:

##  [1] "package"          "license"          "links_to"        
##  [4] "vignette_format"  "vignette_builder" "downloads"       
##  [7] "links_from"       "first_date"       "pub_date"        
## [10] "nrel"

License

Downloads

## Warning in loop_apply(n, do.ply): Removed 2 rows containing non-finite
## values (stat_density).

## Warning in loop_apply(n, do.ply): Removed 1 rows containing non-finite
## values (stat_density).

## Warning in loop_apply(n, do.ply): Removed 1 rows containing non-finite
## values (stat_density).

## Warning in loop_apply(n, do.ply): Removed 1 rows containing non-finite
## values (stat_density).

## Warning in loop_apply(n, do.ply): Removed 1 rows containing non-finite
## values (stat_density).

## Warning in loop_apply(n, do.ply): Removed 1 rows containing non-finite
## values (stat_density).

## Warning in loop_apply(n, do.ply): Removed 1 rows containing non-finite
## values (stat_density).

## Source: local data frame [5 x 3]
## 
##   upstream_repo no_dld yes_dld
## 1     BitBucket      0      23
## 2         Email      0      13
## 3        GitHub      1     825
## 4    None/Other      1    5431
## 5       R-Forge      0     257

I find it hard to believe that organic human-driven downloads would hit essentially every single package on CRAN within a month. Are there automated systems that, e.g., download CRAN in its entirety as a matter of policy?

Vignettes

Number of releases

Date of first and most recent CRAN version

Practices by date of first recent CRAN version

Left: various approaches to testing against date of most recent version.

Right: Frequency of testing for packages that provide an upstream repo (almost all are GitHub) vs those that do not. A proportional bar chart would be nice ... but don't have yet. Having such a repo appears to be strongly correlated with having tests.

Variables I haven't looked at yet

links_to, links_from

Ironholds/practice documentation built on May 7, 2019, 6:41 a.m.