In Party Facts, we only include party information and not the entire original data source.
All original datasets are ignored in version control with the prefix source__ in the local copy.
Files are UTF-8 encoded.
Import of data into the Party Facts database is done by the project maintainers.
For a detailed instruction how to contribute your dataset, see our import guide.
Most of the data preparation is done in R. Some scripts may require additional packages. We have a strong preference for RStudio R packages (esp. tidyverse). run-all.R includes information about used packages.
Please use the tidyverse R Style Guide when writing code in R.
Party Facts import requires ISO3 country codes. country.csv includes the respective information. Recode country names with the R-package countrycode
# Example country recoding from Marpor import
marpor <- marpor %>%
mutate(country = countrycode(countryname, "country.name", "iso3c",
custom_match = c(`Northern Ireland` = "NIR")))
if(any(is.na(marpor$country))) {
warning("Country name clean-up needed")
}Template for dataset readme.md files
# Dataset name
## Source(s)
+ (...) reference in [APSA style](https://connect.apsanet.org/stylemanual/references/)
## Credits (optional)
+ (...) name -- institution -- year(s)
## Import
+ (...) short description of data preparation pipeline
## Linking status (optional)
+ (...) e.g. "completed linking parties > 5% share (staggered import)""
## Comments (optional)
+ (...)
## Todo later (optional)
+ (...)
---
_add figure for dataset if feasible (see examples ParlGov and V-Party)_
