Ways to add data
Every table in Mycorr starts in one of four ways: you upload a file, paste rows, import a table that already lives in Mycorr, or read one through a connection to a database, a bucket or Google Sheets. You can also start from an empty table and type.
Where to start
Open a model, and with nothing selected on the canvas, the action bar at the bottom offers two ways in:
- New Table makes an empty table on the canvas, ready to type into.
- Import opens a menu headed Import from, with three sources: File, Mycorr and, for the model's owners, Connections. Each one opens the Import tab of the panel on the left, on that source.
You can also skip the menu: drag a file from your computer onto the canvas, or paste rows copied from a spreadsheet onto an empty part of it.
| Source | What it brings in | Read more |
|---|---|---|
| File | CSV, TSV, Parquet and Excel files from your computer | Upload a file |
| Paste | Rows copied from a spreadsheet or a text file | Paste data |
| Mycorr | Tables from your other models, datasets, recipes and the public catalog | Import a Mycorr table |
| Connections | Tables from Postgres, Supabase, Google Sheets, Google Cloud Storage and local SQLite files | Connections |
Owned, imported and connected tables
Where a table comes from decides who can change it, and where you change it.
| Owned | Imported | Connected | |
|---|---|---|---|
| Comes from | New Table, an upload or a paste | Another model, a dataset or the catalog | A database, a bucket, a sheet or a SQLite file |
| Where its data lives | In this model | In its home model | At the source |
| Editable here | Yes, by anyone who can edit the model | No, read-only | No, read-only |
| How it changes | You edit it | Its owners release a new version; you take it with Sync to latest release | You change the source, then sync |
| Which version you see | The latest | The release you imported | The latest sync |
An owned table belongs to the model: its rows live there, and every collaborator works on them live. An imported table is a reference, not a copy: your model reads the release its owners published, and stays on that release until you choose to move. A connected table is a copy of what the source held at the last sync, kept read-only so the next sync never overwrites someone's edit.
Who can add tables
| You want to | You need |
|---|---|
| Create an empty table, upload a file, paste rows or import a Mycorr table | To edit the model: a collaborator or an owner |
| Import a table from a connection | To own the model: be a manager of its space |
A collaborator's Import menu has no Connections: a connected table syncs with the credentials of the person who connected it, so connecting one is the owner's. If you can't bring tables in at all, for example as a viewer, there is no Import tab, and a file dragged over the canvas says why: "Only people who can edit this model can bring tables in."
What happens next
Whichever way it arrives, a table is an ordinary table from then on. You can link it to other tables, checkpoint and release it, describe its columns, share it, and add it to a dataset. Its storage counts against the organization that owns the model it lives in; see Usage and limits.