Skip to content

Explore and export

The overview of a dataset

  1. Four figures: rows, columns, duplicates, missing values.
  2. The table, page by page.
  3. Columns: the type and the missing values of each.
  4. Export to a file or a database.
  5. Generate the visualizations.
  6. The pages of the table.
Element Role
Title full path of the dataset: folder, then file name (or engine and database, for a table)
Chips format, size, loading state: at the end of the path, or under it when it fills the line
Four figures rows, columns (numeric and text), duplicates, missing values
Export, Change dataset, Generate the visualizations on the line of the figures, on the right
Table all the rows, page by page
Under the table rows per page (50, 100 or 200) on the left, pages in the centre (or a number to type), rows shown on the right
Columns (side menu, under Overview) opens or closes the columns panel: type (number or text) and missing values of each

Empty cells are written empty. Numbers are aligned right.

Export

Export (overview) writes the dataset on screen, with the loaded rows (a table: its first 10,000 rows), in another format or another database.

To a file: choose the format (CSV, TSV, Excel, Parquet, JSON or YAML), the file name and, for CSV, the separator (comma or semicolon). Then:

  • Download to this computer;
  • or Save on the server, in uploads/: the file then appears in “Choose data”; a name already taken gets “-1”.

To a database: choose the engine and the connection (filled with the last one of the workspace for each engine), the table name and what to do if it already exists: overwrite nothing (the export stops and says so), replace the table or append the rows. The “Columns created” table shows the name and the SQL type each column will get: lower case, no accents, _ instead of spaces. For SQLite, choose a database of the SQLite folder of the server, or New database… and its name (export.db).