Filter by value¶

Every column header carries a small funnel, just to the left of its sort arrows. Click it and a popup lists that column's most common values with their counts, each with a checkbox: tick the ones to keep, press Apply. Clicking the funnel again closes the popup.

It is the fast way to answer "show me just the rows where this column says X", without opening a dialog or typing the value out.

The same filter, a faster door¶

This writes exactly the same per-column allow-set as the Column Filter dialog. It is not a second kind of filter and there is no second code path: the filter chips, the filtered row count, the status bar, the sequential row numbers and every export of the filtered view all behave as they always did.

Right-click a column header → Filter values... opens this same popup, so the funnel and the right-click give one filter. The window version, Columns → Filter by value or shape..., has the same Values / Shapes switch plus a column picker and a Find field: reach for it when you want more room or a column that is scrolled out of view.

What the list shows¶

The 50 most common values, ordered by count. When the column holds more distinct values than that, the popup adds a search box and a line saying how many are not shown.

On a database tab that holds only part of its table, the list comes from the database: the 50 most common values over the whole table, and the search box searches every value there; see Analyses on live databases. A large-file tab does the same over its whole file, in the background, and the ticked values filter every row of the file.

Searching queries the whole column, not the 50 on screen. A value that ranks nine hundredth is still findable by typing part of it. That full scan is exactly why the box appears only when it is needed: opening the popup on an ordinary column stays cheap, and you pay for the deep search only when you ask for one.

All, None, Apply, Clear¶

All and None tick or untick everything currently listed, so they follow the search box rather than ignoring it. Searching for 2024 and pressing All ticks the matching values and leaves the rest alone.
Apply keeps only the ticked values.
Clear removes this column's filter and shows every row again.
Cancel closes the popup without touching the filter.

Two selections deliberately mean "no filter" rather than a filter:

Ticking everything would hide nothing, yet the column would still show as filtered in the chip row and the header. A filter that does nothing but claims to be there is worse than no filter, so it clears.
Ticking nothing would hide every row and leave you looking at an empty table with no obvious way back. That clears too.

Reading the funnel¶

Funnel	Meaning
Grey	The column has no filter
Accent colour	The column is filtered, the pointer is over the funnel, or its popup is open

The small accent dot beside the column name means the same thing. It stays because it is visible on columns too narrow to show the funnel, and because it is easier to spot when scanning a wide table sideways.

The chip row¶

Applying a filter adds a removable chip above the grid, in the same row as every other filter narrowing the view; see The filter chip row.

Partly loaded files¶

On a capped file the counts describe the rows that are loaded, not the whole file, and the popup says so above the list. Reporting a window count as a file count is the kind of quiet inaccuracy this app works to avoid, so the note is not optional.

Shapes instead of values¶

A Values / Shapes switch at the top of the popup swaps the list from values to shapes: what the values look like with the specifics taken out (digits become 9, capitals A, other letters a). Ticking shapes keeps every value that has a ticked shape, which is the fast way to spot the handful of postcodes or IDs typed in the wrong format. See Value Shapes.