Skip to content

Column Statistics

Column statistics summarize the values in one column of a Table view, for example to check that an ID column has no empty cells or to see the range of a price column.

Use either of these:

  • Right-click a column header, or select its ⋮ button, and select Show stats.
  • Right-click a row and select Stats for followed by the column name.

A dialog titled Column: followed by the column name opens, with the column’s type next to the title. While Vaste reads the column, it shows Computing statistics….

Value Meaning
Count Cells that are not empty
Distinct Different values among those cells. Above 100,000, the dialog shows > 100,000.
Nulls Empty cells, including cells missing from rows with too few fields
Min, Max The smallest and largest value
Avg The average, for number columns only
Sample values: The first 20 different values, in the order they appear in the file

A cell that holds only spaces is not empty. The sample values are the first values Vaste meets, not the most common ones.

The type shown next to the title, the same type as the glyph in the column header, decides how Vaste compares values. Vaste picks the type from the first 1,000 data rows.

  • Number columns (integer and float): Min, Max, and Avg use the values that are numbers. If the column holds text further down the file, those cells count in Count but not in Min, Max, or Avg. Whole numbers with 16 or more digits can show rounded.
  • Other columns: Min and Max compare text by character code, so 10 comes before 9, and uppercase letters before lowercase. There is no Avg.

A column of 0 and 1 values counts as a number column, so it gets an Avg.

Statistics always cover the whole column, whatever the current filter or sort. On a large file, reading the column can take a while. The dialog has no progress bar and no cancel button, and closing it does not stop the work.

Vaste keeps the results while the file stays open in Table view, so opening the same column’s statistics again is instant. The column header’s tooltip then also shows the null and distinct counts. Switching views, reloading the file, or changing the CSV dialect clears the results.