RAND.NV in Collabora Office: Use Random Values with Controlled Recalculation
Understand Collabora RAND.NV recalculation triggers, build a scenario worksheet, and preserve a random draw without mistaking nonvolatile for permanent.
Verification status: See the evidence and test scope described in this guide. How formula examples are checked
Quick answer: =RAND.NV() returns a random number between 0 and 1 that does not normally change merely because unrelated input changes. It is still recalculated when the file is opened or a hard recalculation is requested.
Syntax and the meaning of nonvolatile
=RAND.NV()
There are no arguments. Keep the parentheses. The dedicated Collabora section documents these triggers:
| Event | Documented behavior |
|---|---|
| Initial formula input | Generate a random value |
| New unrelated input event | Do not recalculate this nonvolatile function |
| Ordinary F9 | Do not recalculate, except when the cursor is on the formula cell |
| F9 with cursor on the formula cell | Recalculate |
| Recalculate Hard | Recalculate |
| Opening the file | Recalculate |
The help gives the hard-recalculation shortcut as Shift plus the platform’s Command/Ctrl plus F9. Menu and keyboard behavior can vary with platform, so use the named command when checking a deployment.
Nonvolatile does not mean permanently frozen. It also does not mean that a new draw must be different from the previous draw; random draws can repeat.
Example: a one-cell scenario driver
Enter =RAND.NV() in A2. In B2 enter:
=100+20*A2
B2 maps the random input to a scenario value between 100 and 120 under ordinary interval arithmetic. No exact random result is promised. If A2 happened to be 0.25, the independently calculated B2 would be 105; if it were 0.75, B2 would be 115.
The example uses hypothetical input values to make the arithmetic reproducible. It does not pretend that the actual function generated 0.25 or 0.75, nor does it establish a statistical distribution from the vendor’s short description.
For a deterministic check, temporarily use a separate control cell D2 containing 0.25, and compute =100+20*D2 in E2. Keep the random formula in A2 intact so the test data and the live draw are clearly separate.
Verify behavior without requiring a changed number
A practical application check is to record the displayed A2 value, change an unrelated cell, and compare. Then select A2 and invoke F9, or use Recalculate Hard. The help predicts when evaluation occurs, but a matching displayed result after recalculation does not prove no recalculation occurred: display rounding or a repeated draw can conceal the change.
Do not rely on a long list of identical formatted values as a test of randomness. Show additional decimal places when inspecting output, and distinguish formula evaluation from the appearance of a particular number.
These are suggested checks for your installed application; no such Collabora execution was performed for this guide.
Advanced example: compare alternatives with one shared draw
Put costs 100 and 160 in F2:F3 and sensitivities 20 and 10 in G2:G3. With one random driver in A2, use =F2+G2*$A$2 in H2 and fill down.
For the independent control A2=0.25, the two results are 105 and 162.5. Their difference is 57.5. Both scenarios use the same underlying input, so the comparison isolates the different base costs and sensitivities.
Putting a separate RAND.NV call in each row asks a different question because the alternatives no longer share a common draw. Choose intentionally, and retain the inputs if the comparison must be audited later.
ROUND can simplify a displayed scenario result, while SUMPRODUCT can combine a fixed set of weights and scenario values. Neither operation freezes the underlying random driver.
Preserve a result when reproducibility matters
An archived calculation should store the numeric driver used, not only its RAND.NV formula, because opening the file is a documented recalculation trigger. Copying a result into a values-only snapshot is a separate step from making the function nonvolatile. Confirm the copied cell contains a value rather than another formula before treating it as fixed.
Record the scenario formulas and assumptions alongside that snapshot. This function has no documented seed argument, so a saved seed-based replay workflow is not established by its syntax.
Limitations and troubleshooting
The dedicated help says “between 0 and 1” but does not specify endpoint inclusion, a probability distribution, the random-number generator, or statistical quality guarantees. This guide therefore does not promise that 1 is excluded, that draws are independent, or that the function is suitable for cryptography, regulated sampling, or a particular simulation accuracy requirement.
If values change on reopening, that matches the documented behavior. If they change during an explicit hard recalculation, that is also expected. RAND is the related ordinary random function, but its recalculation rules must not be substituted for RAND.NV’s rules.
The vendor lists namespace ORG.LIBREOFFICE.RAND.NV and says the function is outside OpenDocument 1.3 OpenFormula. This identifies an extension; it does not certify other products or platforms.
Availability and verification
The exact function is documented in Collabora Office 26.04 desktop help, retrieved and reviewed on October 17, 2026. This establishes a documentation scope, not the first release that introduced the function or the last version in which it works; those versions and any retirement date remain unestablished. Collabora Online and mobile support are not inferred from desktop help, nor is behavior in another spreadsheet product.
The examples are documentation-based expectations with independent calculations. They have not been executed in Collabora Office. No application screenshot or native error/coercion certification is implied. Formulas use English function names and semicolon argument separators.
References
- Collabora Office 26.04: RAND.NV (checked October 17, 2026).
- Editorial testing standards.