Before any pivot, chart or formula deserves trust, the table underneath it needs to be clean: sorted sensibly, filtered to what matters, free of repeated rows. RixSheets bundles the three everyday tools for this — multi-level sort, auto filter and duplicate removal — alongside find and replace for the typos that always exist. Nothing here is glamorous; all of it is what separates a spreadsheet people believe from one they quietly re-check by hand.
Sort on multiple levels
A single sort rarely answers anything. The useful sorts are compound: by region, then by amount descending; by category, then by date; by status, then by priority.
- Click any cell in the table and open the sort dialog.
- Add your first level — for example, Region, ascending.
- Add a second level — Amount, descending — so the biggest number in each region sits on top.
- Add more levels as needed; RixSheets applies them in order.
- Confirm and review. Repeat sorts are safe: sorting never changes the values, only their order.
Sort before you subtotal, before you present, and before you hand the file to anyone who will read it top to bottom.
Filter with auto filter
Auto filter puts a dropdown on every column header:
- Tick the exact values you want to see, or untick the ones you do not.
- Use the search box inside the dropdown to find a value in a long list.
- Apply number conditions — greater than, between, top items — on numeric columns.
- Apply date conditions on date columns, such as this month or last quarter.
Filtered rows are hidden, not deleted, which makes filtering safe for exploration. Clear the filter to bring everything back. Filters combine across columns: region equals North, amount above 1,000, date this quarter — three dropdowns, one focused view.
Two cautions keep filtering from biting you later. First, a filter is invisible to the next reader: someone opening the file sees fewer rows and may not notice why, so clear filters before saving a file you plan to share, or announce the filter in a note. Second, copying a filtered range copies the hidden rows too unless you copy only the visible selection — a classic source of surprise duplicates downstream, and one more reason the duplicate removal tool below exists.
Find and replace across the table
Recurring typos — "Noth", "SOUTH " with a trailing space, "Acme Inc" versus "Acme Incorporated" — poison groupings and lookups. Open find and replace, search for the bad value, replace with the good one, and let it sweep the sheet in one pass. Do this before you build pivots or joins, because every variant becomes its own group.
Remove duplicate rows
- Click inside the table and open the duplicate removal tool.
- Select the columns that define a duplicate — usually an order number, invoice number or customer plus date combination.
- Review the tool's read of the key columns; include fewer columns only if a shorter key is genuinely correct.
- Run the removal. RixSheets keeps one row per key and reports what was dropped.
- Spot-check the survivors and re-run any formulas that reference the table.
Duplicate removal is deliberately literal: it keeps the first occurrence it meets. Sort first if you want the surviving row to be the most recent or the largest.
Choosing the key columns wisely
The one real decision in duplicate removal is which columns define "the same row", and the wrong choice either deletes too much or too little:
| Key choice | Effect | When it is right |
|---|---|---|
| Order or invoice number | One row per document | The common case — documents are unique |
| Customer + date | One row per customer per day | Contact logs, visit reports |
| Email only | One row per contact | Mailing list imports |
| Every column | Only exact clones removed | When you fear losing data |
If two rows share an order number but disagree on the amount, that is not a duplicate — that is a discrepancy, and collapsing it hides a problem someone needs to see. Duplicate removal is for repeats, not for reconciling conflicts; run it after the typos are fixed and before the totals are trusted.
A cleaning routine that works
| Step | Tool | Why |
|---|---|---|
| 1. Fix typos | Find and replace | Variants split groups and break lookups |
| 2. Remove repeats | Duplicate removal | Double rows inflate every total |
| 3. Sort | Sort dialog | Puts related rows together, exposes oddities |
| 4. Filter | Auto filter | Focuses review on the interesting subset |
| 5. Guard the future | Data validation | Stops the same typos returning |
For criteria too complex for the dropdowns — OR conditions across columns, computed criteria — move up to Advanced Filter. And once the table is clean, data validation keeps it that way while pivot tables turn it into answers.
The routine also pays forward: a table that has been through find and replace, duplicate removal and a sort is a table that imports cleanly elsewhere — into a pivot, into another system via CSV export, or into a report that gets printed. Clean data is the one spreadsheet deliverable that every downstream tool thanks you for.
What to explore next
- Advanced Filter — multi-condition filtering beyond the dropdowns.
- Data validation — prevent bad entries at the source.
- Pivot tables in RixSheets — summarize the table you just cleaned.