Description
Correlation under nonzero linear rescaling; a collapsed variable returns an undefined flag and zero placeholder.
A statistical term depends on its data summaries, weighting convention, centering, denominator, and boundary rules. A contribution, derivative, domain check, or converted summary answers a narrower question than a complete estimator. Those conventions must match before terms are combined or compared.
Correlation under nonzero linear rescaling; a collapsed variable returns an undefined flag and zero placeholder.
When to use this tool
Use Correlation Sign Rescale Policy when you need this specific summary conversion, contribution, domain rule, or sensitivity term with matching statistical conventions.
How to use the tool and how it works
- Enter correlation, firstScale, secondScale. Use the units and conventions below.
- Run the tool. Correct any format, range, or cross-field input error before interpreting the output.
- Read defined, transformed. Inspect defined before interpreting the associated values.
Return transformed=correlation*sign(firstScale)*sign(secondScale) when both scales are nonzero, otherwise zero with defined false. Require a bounded correlation.1
| Input field | Meaning and units | Accepted input |
|---|---|---|
| correlation | Correlation | Required. Finite number. |
| firstScale | First Scale | Required. Finite number. |
| secondScale | Second Scale | Required. Finite number. |
| Output field | Interpretation | Returned shape |
|---|---|---|
| defined | Whether the stated denominator or quantity is defined; inspect before interpreting a placeholder. | Required. Boolean. |
| transformed | Transformed | Required. Finite number. |
{
"input": {
"correlation": 0.6,
"firstScale": -2,
"secondScale": 3
},
"output": {
"defined": true,
"transformed": -0.6
}
}Limitations and assumptions
- Zero-scale output is a placeholder, not a meaningful correlation. This assumes affine rescaling of nonconstant variables and does not estimate correlation from data or resolve invalid original summaries.
- The field tables state structural bounds; the mechanics also specify cross-field constraints. Malformed inputs and unsupported domains cause errors. A successful false flag, empty branch value, or zero placeholder has the meaning stated above and is distinct from an input error.
- Numeric calculations use binary floating point unless exact decimal-integer arithmetic is explicitly stated. No hidden measurement tolerance or decimal rounding is added; nonfinite or out-of-contract results are rejected. Signed zero is exported as zero.
Alternative or Complementary methods
Use a complete estimator or model when a contribution, sensitivity, or domain check is insufficient. Verify the weight semantics, centering, denominator, and sampling assumptions before combining summaries.
References
Similar or alternative tools
- Correlation Squared Sign Loss Label
Return squared=correlation² and the original correlation's sign, requiring correlation within minus one through one.
- Correlation Attenuation Reliability Factor
Calculate corrected=observed/sqrt(firstReliability*secondReliability) and return withinCorrelationRange for magnitude at most one.