DATA ANALYSIS PACKAGE SYMBOL
evaluates expr while propagating standard errors of DV and NDV values by linearizing the functions in the expression.
- DVEvaluate uses linearizations to propagate uncertainties, requiring that the partial derivatives can be calculated for the expression.
- For non-correlated errors, the standard error for an expression expr with NDV/DV values x1, x2, ... is calculated as seexpr=.
- For highly non-linear expressions, the linearizations are not valid and MCEvaluate or MonteCarloSimulation should be used instead.
- The following options can be given:
|CorrelatedErrors||False||CorrelatedErrors specifies whether to assume that all random errors are perfectly correlated (True) or completely uncorrelated (False).|
|AssumeND||True||AssumeND specifies whether to assume that the resulting distribution is a normal distribution (True) or not (False).|
Propagate uncertainties for a simple expression:
Uncertainties of values in DataTable objects are automatically propagated when using Query functions: