I have a CSV file called startdates.csv which I want to provide to colleagues as part of an R package called tabulations. The individual steps work during development but when the package is installed and loaded, the data.frame columns are concatenated. Here are the details:
As per Hadley Wickham's excellent book and this very helpful summary, I put startdates.csv in data-raw/, and ran the following script to save it as data/startdates.rda:
startdates <- read.csv("data-raw/startdates.csv")
devtools::use_data(startdates, overwrite=T)
which works as expected:
> source("data-raw/get_tabs.r")
Saving startdates as startdates.rda to /home/matthewroberts7/ieTabs/tabulations/data
> startdates
Dataset Observations Potential_start_date_fields Earliest Latest Missing
1 jeval 6445712 5 30NOV2000 21NOV**** 0.04%
2 jhist 7058169 5 12MAY2001 17SEP**** 2.90%
3 ueval 390783 5 29APR2013 29JUN2016 0.00%
During development, all is well:
> setwd("tabulations")
> devtools::load_all()
Loading tabulations
> tabulations::startdates
Dataset Observations Potential_start_date_fields Earliest Latest Missing
1 jeval 6445712 5 30NOV2000 21NOV**** 0.04%
2 jhist 7058169 5 12MAY2001 17SEP**** 2.90%
3 ueval 390783 5 29APR2013 29JUN2016 0.00%
But when the package is installed it's as if the original csv has been saved as a single string per row:
> setwd("..")
> document("tabulations")
Updating tabulations documentation
Loading tabulations
> install("tabulations")
Installing tabulations
...
> reload(inst("tabulations"))
Reloading installed tabulations
> tabulations::startdates
Dataset.Observations.Potential_start_date_fields.Earliest.Latest.Missing
1 jeval,6445712,5,30NOV2000,21NOV****,0.04%
2 jhist,7058169,5,12MAY2001,17SEP****,2.90%
3 ueval,390783,5,29APR2013,29JUN2016,0.00%
Can anyone explain what's going on here? Many thanks.
EDIT: R v3.4.2, devtools v1.13.4
