reparametrize {dsfa}

R Documentation

reparametrize

Description

Transforms the given inputs to the parameters and the first three moments of the corresponding distribution. For the normal-halfnormal distribution the parametrization of the classical stochastic frontier as well as the skew-normal and centred skew-normal specification ar provided. For the normal-exponential an the specification via \lambda is available.

Usage

reparametrize(
  mu = NULL,
  sigma_v = NULL,
  sigma_u = NULL,
  s = NULL,
  lambda = NULL,
  par_u = NULL,
  mean = NULL,
  sd = NULL,
  skew = NULL,
  family = NULL
)

Arguments

mu	vector of \mu
sigma_v	vector of \sigma_V. Must be positive.
sigma_u	vector of \sigma_U. Must be positive.
s	s=-1 for production and s=1 for cost function.
lambda	vector of \lambda. Must be positive.
par_u	vector of \sigma_U or \lambda. Must be positive.
mean	vector of mean of \mathcal{E}
sd	vector of standard deviation of \mathcal{E}. Must be positive.
skew	vector of skewness of \mathcal{E}.
family	normhnorm for normal-halfnormal and normexp for normal-exponential distribution.

Details

The following input combinations are allowed for the normal-halfnormal distribution

• mu, sigma_v, sigma_u, s

• mean, sd, skew, family="normhnorm" with optional s \qquad,

while for the normal-exponential distribution the feasible inputs are

• mu, sigma_v, lambda, s

• mean, sd, skew, family="normexp" with optional s \qquad.

Other input combinations are not feasible.

Value

Returns a data.frame with the parameter values for all specification.

References

• Kumbhakar SC, Wang H, Horncastle AP (2015). A practitioner's guide to stochastic frontier analysis using Stata. Cambridge University Press.

• Azzalini A (2013). The skew-normal and related families, volume 3. Cambridge University Press.

Examples

#Normal-halfnormal distribution
para<-reparametrize(mu=1, sigma_v=2, sigma_u=3,s=-1)
reparametrize(mean=para$mean, sd=para$sd, skew=para$skew, family="normhnorm")

#Normal-exponential distribution
para<-reparametrize(mu=1, sigma_v=2, lambda=1/3,s=-1)
reparametrize(mean=para$mean, sd=para$sd, skew=para$skew, family="normexp")

---