Trace Formula Precedents and Dependents in ONLYOFFICE

Quick answer

Select a formula cell and choose Formula → Trace Precedents to inspect its inputs. Select an input and choose Trace Dependents to inspect formulas using it. Arrows reveal dependencies, not whether the business calculation is correct. A wrong formula can produce an ordinary number with no error message.

Scope and a safe starting point

This guide targets the English free Desktop Editors 10.0.0 application, not a paid Docs server or DocSpace account. The desktop changelog identifies 10.0.0, and Desktop Editors is free for personal and professional use. The Formula tab help includes the desktop interface and tracer tools. The official tracing explanation dates their introduction to 7.5; earlier installations may lack them.

Create a new local practice workbook named Trace-Practice.xlsx, with a sheet named Model. Do not use a real financial model. Record the saved location and actual application version. The following numbers are synthetic, and the outputs are independently calculated expectations rather than observed ONLYOFFICE results.

Build a model with one deliberate mistake

Enter the input labels in column A and values in column B:

Cell	Label in column A	Value in column B
B2	Units	8
B3	Unit price	15
B4	Discount rate	10%
B5	Fixed cost	25

Enter these labels in D2:D5 and formulas in E2:E5. Type B4 as 10%, not 10.

Row	Column D	Formula in column E	Initial result
2	Gross amount	=B2*B3	120
3	Discount amount	=E2*B4	12
4	Correct net	=E2-E3-B5	83
5	Suspect net	=E2-B4-B5	94.9

E5 subtracts a rate, 0.1, where the model needs a discount amount of 12. It is syntactically valid but represents the wrong calculation. Its difference from the correct net is 11.9. Formatting E5 as currency would not repair that mistake.

Trace the two output cells

1. Select E4, then choose Formula → Trace Precedents. Compare the referenced cells with the formula bar: the direct inputs are E2, E3 and B5.
2. Choose Formula → Remove arrows to clear the display before the next comparison.
3. Select E5 and trace its precedents. Its formula directly references E2, B4 and B5. It does not reference the calculated discount in E3.
4. Read each referenced cell’s label and unit. B4 is a rate; E3 is an amount. This semantic difference explains the error even though both cells contain numbers.
5. Inspect E3 separately with Trace Precedents. Its formula depends on E2 and B4, which is the intended multiplication of gross amount by discount rate.

These checks follow the documented selection-and-trace workflow. They do not require a particular arrow route, pixel position, or separate arrow for every cell in a larger range. Judge the referenced addresses and formula together. If the display is confusing, clear the arrows and inspect one result at a time.

Trace an input’s immediate consumers

Clear the arrows, select B4, and choose Formula → Trace Dependents. Before the repair, E3 and E5 directly use B4. E4 also depends on the rate indirectly through E3; inspecting one level is not proof that you have inventoried every downstream consequence.

To follow that chain deliberately, inspect E3’s dependents separately. This is a useful distinction when a model has several calculation stages: an input can affect a result without appearing literally in the result’s formula.

Change E5 to =E2-E3-B5. Both net cells should now show 83. Clear and reapply the traces after editing, rather than assuming already displayed arrows have updated. B4 should no longer be a direct reference in E5; E3 should be.

Test the repaired model rather than trusting the arrows

Perform these edits in sequence and inspect all four result cells:

Test state	E2 gross	E3 discount	E4 net	E5 repaired net
B2=8, B4=10%	120	12	83	83
Change B2 to 10	150	15	110	110
Keep B2=10; change B4 to 0%	150	0	125	125
Restore B2=8 and B4=10%	120	12	83	83

The zero-rate case is important: the original bad formula would happen to agree with the correct formula at zero. Passing that one case alone would miss the defect. The nonzero case and dependency inspection distinguish the two calculations.

If a result differs, inspect the cell’s actual formula and input type. Do not replace a result with its expected number. For merely watching distant values while changing inputs, use the separate Watch Window guide; it does not replace this dependency audit.

Clear the visual aids and preserve the repaired formulas

Use Formula → Remove arrows when finished. Its menu also distinguishes precedent and dependent arrows. Removing the visual aids should not delete the source cells or formulas; inspect E4 and E5 afterward.

Save the local XLSX, close it, and reopen the same path. Verify the restored inputs and both formulas yielding 83. Reapply traces if needed. This guide does not establish whether tracer displays persist through reopening, so saved arrow visibility is not an acceptance criterion.

The exercise is limited to ordinary direct references on one worksheet. External workbooks, cross-sheet tracer navigation, INDIRECT/OFFSET-generated references, named-formula expansion, circular references and exported-file behavior need separate checks. No tracer display proves that the underlying data or business rule is correct.

Verification: Official pages were read on 21 October 2026. Arithmetic and the explicit dependency graph were checked independently. No application was launched for this guide; arrow rendering, interaction and save/reopen remain untested. No screenshot is presented as native evidence.