User Guide
Mapper is a generic tool to map different datasets to a pre-established standard. It can be used to implement data interoperability solutions. The Mapper allows mapping columns and column values from files that contain data to be harmonised. Each file to be mapped contains data from one entity of the standard. The files are csv files and can use practically any standard separator besides comma.
Find more in this example.
The application has three main functions:
- Map attributes of entities
- Map attribute values
- Save the map
The standard is expressed mainly with two files. One for the entities and attributes per entity and another for the attribute values. You can find an example here.
The output of the Mapper is a zip files containing JSON files of the map and column lists from the mapped files.
In order to map your datasets you need to know the semantic behind the standard. You can download the standard by clicking the button "Download Standard" as in the figure below:
Each entity in the standard corresponds to a data file that contains data to be mapped. In the standard file, the fact entity is the first entity in the columns "Entity". In the web interface of the Mapper, the fact entity is the first entity in the entity list.
The fact entity or central entity is the first entity in the dropdown list containing the entities. For instance, the figure below show a standard for biobank data based in MIABIS standard. The fact entity is Sample.
The rest of entities in the standard are the rest of the list following the fact entity. In the example above, the rest of entities to be mapped are: Sample Collection, Study and Contact Information.
The attributes to be mapped are those from the standard defined in the web service. Nevertheless, the files to be mapped can contain any number of columns in any order.
The datasets are the cvs files containing data to be mapped according to the standard defined by the Mapper web service that you are using. In our biobank federation example, the biobanks will expose sample data for sharing. The biobanks export cvs files with data regarding: sample, sample collection, study and contact person. Those files can differ a lot from one biobank to another. Each biobank maps its files to allow them to be processed by a federation framework (MIABIS Federation that allows to query all the data from all the biobanks in the federation.
To map attributes from one entity (columns in a file):
- Select the Entity in the Entity list (figure above)
- Chose the file containing the data for the selected entity
- Map attributes by clicking first in the "Standard Attributes" table and then in the "Local Attributes" table. The standard name for the attribute will appear in the third column of the "Local Attributes" table.
Repeat 3 until all possible columns are mapped.
- Save the map for the selected entity
