Update a dataset
Updating a dataset can be done manually on the dataset edition page or can be automated with periodic processing.
Before performing a manual update, check your data schema. For example, if a visualization uses your dataset, verify that the columns used by the visualization are still present in the new file you are importing.
The update is done using the Update action button on the dataset edit page you want to update.
You need to choose the new file to load on your computer and upload your dataset on Data Fair.
The file will be submitted to 6 processing steps:
- The Loading, which represents the progress bar.
- The Conversion to a format used by the platform.
- The Analysis, which will determine the schema of the dataset.
- Indexing, which will allow quick access to the data in the file.
- The "Enrichment, which will take into account the new values of the dataset and will perform a new enrichment according to the extensions added for the dataset.
- The Finalization, which corresponds to the last processing before the dataset is available.
When the finalization is complete, the dataset goes into "draft".
Draft mode allows to check the structure of the schema to add concepts to the new columns and to consult the first 100 lines of the new file.
New columns will be displayed in red.
You can cancel the update if the schema does not match what you want.
After performing the schema verification and giving, the Validate draft button allows you to launch the last stage of the update.
The file will again be subjected to the 6 processing steps.
When finalization is complete, the dataset goes to "available" state.
It can then be edited, enriched and used in visualizations.