Check how a file was imported, and change a column’s type
See rows the import left out and columns it may have misread, and import again with the right type.
For the complete AI-readable content index, see /llms.txt.
Datasets · 2 min read
When a CSV or Excel file (or a Google Sheet) is imported, DashboardGPT works out each column’s type from its values. The Profile tab says what it could not do cleanly: rows it could not read and left out (with their line numbers), columns it had to name because the header was empty, repeated or missing, and columns whose values suggest the wrong type.
Typical misreads: a zip code, phone number or SKU read as a number (it can be added up and loses leading zeros), a year read as a quantity, amounts written like “$1,200” read as text so they can’t be summed, and dates in an unusual format read as text so they can’t be grouped by month.
To change a type, choose “Change to …” on the Profile tab, or the pencil next to a type on the Schema tab, pick the type and choose “Change type and re-import”. The data is imported again from the uploaded file, or the Google Sheet is refreshed, with the column read as the new type. Charts keep the current data until the import finishes.
The dialog lists the charts that use the column and warns about any the new type would stop working, for example a sum of a column changed to text. Values that don’t fit the new type are left empty and counted on the Profile tab.
Related reading
Guides from Learn on the why and when behind this.
Still stuck?
Ask the docs assistant about “Check how a file was imported, and change a column’s type” or anything else in DashboardGPT.