R/schema_conf.R
sourcing_metadata.Rd
The set of function allows to configure which data information should be saved to configuration YAML file when such configuration is sourced directly from database schema.
opt_pull_character(
values = TRUE,
max_uniq_to_pull = 10,
nchar = TRUE,
na_ratio = TRUE,
levels_ratio = TRUE,
...
)
opt_pull_numeric(
values = TRUE,
max_uniq_to_pull = 10,
range = TRUE,
precision = TRUE,
scale = TRUE,
na_ratio = TRUE,
levels_ratio = FALSE,
...
)
opt_pull_integer(
values = TRUE,
max_uniq_to_pull = 10,
range = TRUE,
na_ratio = TRUE,
levels_ratio = FALSE,
...
)
opt_pull_date(range = TRUE, na_ratio = TRUE, levels_ratio = FALSE, ...)
opt_pull_logical(na_ratio = TRUE, levels_ratio = FALSE, ...)
opt_pull_table(nrows = "exact", ...)
Should column unique values be sourced? If so the ones are stored as
an array withing values
parameter.
Pull unique values only when the distinct number of them is less than provided value. The parameter prevents for sourcing large amount of values to configuration file for example when dealing with ids column.
Should maximum number of characters in column be pulled? Is so stored as
nchar
parameter in configuration YAML file.
Should ratio of NA values existing in column be sourced?
Should ratio of unique column values be sourced?
Other parameters defining column metadata source. Currently unsupported.
Should column range be sourced? Is so stored as range
parameter in
configuration YAML file.
Currently unused.
Currently unused.
Should number of original columns be sourced? When 'exact' stored as a nrows
parameter for each table in YAML configuration file. When 'ratio' stored as a fraction of original columns
(based on all tables) and saved as nrows
configuration parameter. When 'none'
tables rows information will not be saved.