update_attributes: Update minimum and maximum values of numeric variables in...

In CAPLTER/capeml: CAP LTER package for generating EML

Update minimum and maximum values of numeric variables in existing attribute metadata

Description

A frequent need with long-term, ongoing research is to update existing data. A challenge to that is that we do not want to have to rebuild from scratch the attribute metadata for a data entity at each update. In terms of attribute metadata, definitions, units, etc. are relatively static but what can change are the minimum and maximum values for numeric variables as the observation record grows. We could ascertain the minimum and maximum values for numeric variables then manually update existing attribute metadata but this is tedious, error-prone, and can be time consuming when dealing with many variables. The update_attributes function takes care of this for us by reading the existing attribute metadata for a given data entity and updating those metadata with the minimum and maximum values for said data entity if they have changed in the context of a data refresh.

Usage

update_attributes(entity_name, return_type = "yaml")

Arguments

entity_name	(character) The name of the data entity.
return_type	(character) Quoted designator indicating the value returned as either a attributes template yaml file (return_type = "yaml", the default) or a list of entity attributes (return_type = "attributes") constructed from the data entity. The latter (i.e., return_type = "attributes") is really just for testing the function.

Value

An updated metadata template including refreshed minimum and maximum vales for numeric variables for providing attribute metadata as a yaml file with the file name of the R data object + "_attrs.yaml" in the working directory.

Note

An artifact of the updating process is that empty/unused keys will be omitted from the updated yaml file. For example, empty annotation fields will not be present in the yaml file after the update. This does not affect functionality of the metadata generated (the keys were not used anyway) but the keys would have to be added manually if there was a need for them in the future.

An artifact of the updating process is that the definition element is populated in the updated yaml file. The definition element is required by the EML schema for attributes of type character. In a typical workflow, the definition element is left blank in the metadata yaml file and is automatically populated with a copy of the attributeDefinition element at build time (e.g., create_dataTable). Because update_attributes is calling read_attributes, which is what populates any empty definitions for variables of type character, this is reflected by update_attributes in the updated yaml. This does not have any bearing on the resulting data entity EML, and we can still provide custom definition metadata if desired.

update_attributes will abort the update if an attribute is detected in the data entity but for which there is not metadata in the existing attributes file. This is an indication that the data structure or content has changed sufficiently that a new, blank attributes metadata file should be constructed. Conversely, if an attribute is detected in the existing metadata that is not detected in the data entity, the update will proceed but the attribute and corresponding metadata in the entity but not in the existing metadata file will be stricken from the updated attribute metadata file. In both cases, update_attributes will print to screen the incongruent attributes.

Examples

## Not run: 

 # update attributes file for mycars data object

 mycars <- head(mtcars)

 capeml::update_attributes(entity_name = mycars)


## End(Not run)

CAPLTER/capeml documentation built on April 3, 2024, 11:17 p.m.