- Added a
NEWS.md
file to track changes to the package. - Added
canonical_names()
to create a subjective canonical name for GESIS variables.
- The
gesis_vocabulary_create()
helper function is added to make variable label and value label harmonization easier. - New
vignette(vocabulary)
is developed as a working article to start creating meaningful metadata files. This is at this stage intended for contributors. See Working With Vocabularies.
- The Harmonizing Variable Names vignette/article started for an early standardization of variable names. Thanks for the contribution from Marta.
- In the future,
canonical_names()
may be renamed, and it may receive a new functionality. - The
gesis_vocabulary_create()
and Thegesis_metadata_create()
have better variable names. Thenormalize_names()
function is renamed tolabel_normalize()
. - Internal function is renamed to
class_suggest()
fromsuggest_conversion()
. - The Workflow vignette is open for consultation.
label_normalize()
has unit tests, new documentation, and the inputs parameters are synchronized with the other variable name conventions (in metadata). Open question: what to do with non-standard latin characters? [To be resolved in 0.1.3].canonical_names()
is renamed tolabel_suggest()
and has unit tests.naming functions
is renamed tolabelling functions
with the British spelling.- Three new Eurobarometer sample files to be used in vignettes and unit testing:
ZA7489_sample
,ZA7576_sample
,ZA7562_sample
. These samples contain countries that require special attention (DE
,GB
,CY
,MK
), some metadata, some demograpy variables, some trending (trust) variables and some ad hoc variables. read_example_file()
is a simple wrapper aroundutils::data()
to mimick the data importing workflow withhaven::read_spss()
or other importing functions in vignette examples and unit tests.panel_create()
creates a skeleton panel with a unique panel id made of selected ID variables, such as the individually unique case ID and the doi of the survey.gesis_metadata_create()
can now take either a list of surveys, or a single survey data frame as an input, and returns thefilename
in a column. Two new columns, the filename and the suggested question block is added to the metadata (see Workflow)gesis_metadata_create()
is replaced with the more genericread_surveys()
.- Started continuous integration on Travis-CI.
id_create
tests unicity andpanel_create()
is a wrapper around it.
read_surveys ( my_spss_files, .f= 'read_spss_survey')
now reads in a list of surveys with recording user-specified SPSS missing values.
harmonize_value_labels()
is harmonizing (hopefully) all binary categorical values. The workflow vignette is updated.
- Definition of new class with the constructor
eurobarometer_labelled()
- Preliminary analysis of the vocabulary of 97 SPSS files (corresponding to 97 Eurobarometer waves): standardization of spelling of common (and often abbreviated) words in variable labels, tokenization of variable labels, creation of a doc-term matrix for variable labels, creation of preliminary standardized variable names.
- Serveral functions were deprecated:
concatenate
,unique_id_create
,harmonize_to_numeric
,eurobarometer_labelled
and theworkflow
vignette and the NUTS1 and NUTS2 vocabulary files.
Fully harmonized with retroharmonize 0.1.1.