Validates the metadata and data points against the extraction form and
writes data/<study_id>/ (study.yml, metadata.yml, data.csv,
provenance.csv and datapackage.json; see SPEC.md section 8). All
validation problems are reported at once.
Usage
add_study_data(
path = ".",
study_id,
metadata,
data,
reports = NULL,
provenance = list(channel = "manual_edit"),
overwrite = FALSE,
validate = TRUE
)Arguments
- path
Project root.
- study_id
Study id (see
check_study_id()), e.g."hansen2019".- metadata
Named list matching the
metadatapart of the form. Arrays of objects may be given as data frames.- data
Data frame (or list of objects) matching
data_points. May have zero rows.- reports
Reports (publications) of the study: a list of lists with
doi,pmidandrole(primaryorsecondary). Defaults to the existing reports, elsemetadata$doi/metadata$pmid.- provenance
Where the values came from: either a list of defaults (
extraction_id,person,channel,tool,model,source_sha256,timestamp) applied to every data point, or a data frame with ameasurement_idcolumn (and_metadatafor the metadata). When overwriting, the provenance of data points whose values did not change is kept.- overwrite
Replace an existing study? Defaults to
FALSE.- validate
Validate before writing? Defaults to
TRUE.
Details
This bypasses the submission workflow. Input channels should use
submit_study() and accept_submission(), which call this function and
record where each value came from.