Trend R6 class

Trend R6 class

Trend R6 class

Format

R6Class object.

Value

Object of R6Class with methods for fitting GP model.

Super class

GauPro::GauPro_trend -> GauPro_trend_LM

Public fields

m

Trend parameters

m_lower

m lower bound

m_upper

m upper bound

m_est

Should m be estimated?

b

trend parameter

b_lower

trend lower bounds

b_upper

trend upper bounds

b_est

Should b be estimated?

Methods

Public methods

• trend_LM$new()

• trend_LM$Z()

• trend_LM$dZ_dparams()

• trend_LM$dZ_dx()

• trend_LM$param_optim_start()

• trend_LM$param_optim_start0()

• trend_LM$param_optim_lower()

• trend_LM$param_optim_upper()

• trend_LM$set_params_from_optim()

• trend_LM$clone()

---

Method new()

Initialize trend object

Usage

trend_LM$new(
  D,
  m = rep(0, D),
  m_lower = rep(-Inf, D),
  m_upper = rep(Inf, D),
  m_est = rep(TRUE, D),
  b = 0,
  b_lower = -Inf,
  b_upper = Inf,
  b_est = TRUE
)

Arguments

D

Number of input dimensions of data

m

trend initial parameters

m_lower

trend lower bounds

m_upper

trend upper bounds

m_est

Logical of whether each param should be estimated

b

trend parameter

b_lower

trend lower bounds

b_upper

trend upper bounds

b_est

Should b be estimated?

---

Method Z()

Get trend value for given matrix X

Usage

trend_LM$Z(X, m = self$m, b = self$b, params = NULL)

Arguments

X

matrix of points

m

trend parameters

b

trend parameters (slopes)

params

trend parameters

---

Method dZ_dparams()

Derivative of trend with respect to trend parameters

Usage

trend_LM$dZ_dparams(X, m = self$m_est, b = self$b_est, params = NULL)

Arguments

X

matrix of points

m

trend values

b

trend intercept

params

overrides m

---

Method dZ_dx()

Derivative of trend with respect to X

Usage

trend_LM$dZ_dx(X, m = self$m, params = NULL)

Arguments

X

matrix of points

m

trend values

params

overrides m

---

Method param_optim_start()

Get parameter initial point for optimization

Usage

trend_LM$param_optim_start(
  jitter = FALSE,
  b_est = self$b_est,
  m_est = self$m_est
)

Arguments

jitter

Not used

b_est

If the mean should be estimated.

m_est

If the linear terms should be estimated.

---

Method param_optim_start0()

Get parameter initial point for optimization

Usage

trend_LM$param_optim_start0(
  jitter = FALSE,
  b_est = self$b_est,
  m_est = self$m_est
)

Arguments

jitter

Not used

b_est

If the mean should be estimated.

m_est

If the linear terms should be estimated.

---

Method param_optim_lower()

Get parameter lower bounds for optimization

Usage

trend_LM$param_optim_lower(b_est = self$b_est, m_est = self$m_est)

Arguments

b_est

If the mean should be estimated.

m_est

If the linear terms should be estimated.

---

Method param_optim_upper()

Get parameter upper bounds for optimization

Usage

trend_LM$param_optim_upper(b_est = self$b_est, m_est = self$m_est)

Arguments

b_est

If the mean should be estimated.

m_est

If the linear terms should be estimated.

---

Method set_params_from_optim()

Set parameters after optimization

Usage

trend_LM$set_params_from_optim(optim_out)

Arguments

optim_out

Output from optim

---

Method clone()

The objects of this class are cloneable with this method.

Usage

trend_LM$clone(deep = FALSE)

Arguments

deep

Whether to make a deep clone.

Examples

t1 <- trend_LM$new(D=2)