One dictionary variable, normalised for the index.
Adapters emit one row per variable defined in a dataset’s data dictionary. A
variable is defined once per logical table, even when the physical CSV repeats
across family-role folders (as in SSC).
-
dataset
Dataset id the variable belongs to.
- Type:
str
-
version
Dataset version the variable was read from.
- Type:
str
-
table_name
Logical table the variable belongs to (the CSV stem the user sees).
- Type:
str
-
name
Variable (column) name as it appears in the data files.
- Type:
str
-
qualified_id
Fully qualified id from the dictionary, when one is given (SSC).
- Type:
str
-
definition
The variable’s definition or question text.
- Type:
str
-
field_type
Storage or answer type recorded by the vendor (SPARK).
- Type:
str
-
measurement_scale
Measurement scale such as nominal or ordinal (SSC).
- Type:
str
-
value_coding
Coded answer choices, for example 0 = no / 1 = yes.
- Type:
str
-
notes
Free-text notes from the dictionary.
- Type:
str
-
display_title
Human-facing label for the variable (SSC).
- Type:
str
-
display_hint
Short hint shown alongside the label (SSC).
- Type:
str
-
roles_applicable
Comma-separated family roles the variable applies to (SSC); empty for SPARK.
- Type:
str
-
source_sheet
Dictionary sheet the row was parsed from.
- Type:
str