Skip to contents

This function generates an empirical CDF using partial moment ratios LPM.ratio, and resulting survival, hazard and cumulative hazard functions.

Usage

NNS.CDF(variable, degree = 0, target = NULL, type = "CDF", plot = TRUE)

Arguments

variable

a numeric vector or data.frame of >= 2 variables for joint CDF.

degree

numeric; (degree = 0) (default) is frequency, the empirical distribution of variable; (degree = 1) is area, the share of deviation area below each point, which is not \(P(X \le t)\).

target

numeric; NULL (default) Must lie within support of each variable (one value per column for multivariate variable).

type

options("CDF", "survival", "hazard", "cumulative hazard"); "CDF" (default) Selects type of function to return. For multivariate variable the joint survival is \(P(X_1 > t_1, \ldots, X_d > t_d)\) at degree = 0, by inclusion-exclusion over the column subsets for \(d \le 3\) and as the share of observations above \(t\) in every column for larger \(d\), and the joint upper partial moment Co.UPM_nD at degree > 0, normalised like the CDF. The multivariate cumulative hazard is the transform \(-\log S(t)\) and requires degree = 0; it is not automatically an integral of a scalar multivariate hazard. Multivariate type = "hazard" is unsupported and raises an error: a joint density-over-survival or a coordinate-hazard definition requires a separate estimator, and dividing the joint CDF by survival does not estimate either. At degree > 0 with two or more columns, each target's co-partial moment is normalised by its own total, so the "CDF" and "survival" values are generalised co-partial-moment ratios, not a joint distribution function: they can assign a negative probability to a rectangle. At degree = 0, and in one dimension at any degree, they are proper distribution functions; at positive degree the univariate induced law describes deviation-area balance rather than the frequency law of the original variable.

plot

logical; plots CDF.

Value

Returns:

  • "Function" a data.frame (class NNS.data.frame) containing the observations and the resulting function of the variable.

  • "target.value" value from the target argument.

References

Viole, F. and Nawrocki, D. (2013) "Nonlinear Nonparametric Statistics: Using Partial Moments" (ISBN: 1490523995, 2nd edition: https://ovvo-financial.github.io/NNS/book/)

Viole, F. (2017) "Continuous CDFs and ANOVA with NNS" doi:10.2139/ssrn.3007373

Author

Fred Viole, OVVO Financial Systems

Examples

if (FALSE) { # \dontrun{
set.seed(123)
x <- rnorm(100)
NNS.CDF(x)

## Empirical CDF (degree = 0)
NNS.CDF(x)

## Area-weighted CDF (degree = 1): the share of deviation area below each point
NNS.CDF(x, 1)

## Joint CDF
x <- rnorm(5000) ; y <- rnorm(5000)
A <- cbind(x,y)

NNS.CDF(A, 0)

## Joint CDF with target
NNS.CDF(A, 0, target = rep(0, ncol(A)))
} # }