Text in project_name and city is valid UTF-8 again. The raw CSV mixes Windows-1252 and UTF-8 text; data-raw/data_processing.R now converts the 29 values that were not valid UTF-8 (26 in project_name, 3 in city) from Windows-1252, so accented names such as “SÃO PAULO”, “Lomé” and “Ceará”, typographic dashes and apostrophes come out correctly (#5).
The CSV and the XLSX export are written from the same object. The XLSX file is now inst/extdata/washinvestments.xlsx; the old washinvestments_utf8.xlsx, in which the invalid characters were removed, is gone (#5).
The exact duplicate row for project 53284-001 is removed, so the data have 1,872 records instead of 1,873 (#5).
The region “Nothern Europe” is corrected to “Northern Europe” (#5).