PCPMr R6 Class

This class wraps the PCPM C++ class using R6.

Public fields

N

Number of observations.

J

Number of items.

K

Number of latent factors.

P

Number of predictors.

Q

Loading pattern matrix.

Qb

Regression pattern matrix.

LD

Logical indicating local dependence.

DIF

Logical indicating differential item functioning.

time

Computation time.

MODEL

Model type identifier.

FacSco

Stored factor scores.

Methods

Public methods

• PCPMr$new()

• PCPMr$pcfa()

• PCPMr$getVariable()

• PCPMr$pcpmsummary()

• PCPMr$clone()

---

Method new()

Initializes a new PCPM object.

Usage

PCPMr$new(
  priorF = c(0.1),
  priorA = c(1, 0.1, 1, 0.1, 0, 0.1),
  priorB = c(1, 0.1, 1, 0.1, 0, 0.1)
)

Arguments

priorF

A numeric vector for priorF (default: c(0.1, 1.0, 1.0)). w0

priorA

A numeric vector for priorA (default: c(1.0, 1.0, 1.0, 1.0, 0.0, 1.0)).s0;r0;lam_a;lam_b;m0;c0;

priorB

A numeric vector for priorB (default: c(1.0, 1.0, 1.0, 1.0, 0.0, 1.0)).s0B;r0B;lam_aB;lam_bB;m0B;c0B;

---

Method pcfa()

Performs PCFA on the provided data.

Usage

PCPMr$pcfa(
  Y,
  Q,
  cati = F,
  cati_v = -1,
  reglo = "lasso",
  LD = F,
  regphi = "lasso",
  Kg = 0,
  reg,
  regpsx = "lasso",
  iter = 1000,
  burn = 0,
  std = T,
  cor_fac = 1,
  sign_check = F,
  update = 1000
)

Arguments

Y

A numeric matrix.

Q

An integer matrix.

cati

A logical value, wheter item is categorical.

cati_v

A vector or -1. which item is categorical, -1 means all.

reglo

A string. which prior for loading. 'lasso','horse','ssp'

LD

A logical value.

regphi

A string. which prior for factor correlation. 'none', lasso','horse','ssp','parlasso','parhorse','parssp'

Kg

A numeric. how many general factor, if none, 0

reg

An integer matrix. set for par phi, not used here. default as matrix full with 1.

regpsx

A string. which prior for loading. 'lasso','horse','ssp'

iter

An integer value for the number of iterations.

burn

An integer value for the burn-in period.

std

A logical value indicating whether to standardize.

cor_fac

A numeric vector for correlation factors.

sign_check

A logical value.

update

An integer value for the update frequency (default: 1000).

---

Method getVariable()

get posterial summary

Usage

PCPMr$getVariable(varName)

Arguments

varName

aaa.

---

Method pcpmsummary()

Performs pcpmsummary.

Usage

PCPMr$pcpmsummary(
  med = FALSE,
  start = 0,
  end = -1,
  SL = 0.05,
  sig = T,
  detail = F,
  SignSwitch = F,
  reorder = F
)

Arguments

med

aaa.

start

aaa.

end

aaaa

SL

aaa

sig

aaa

detail

aaa

SignSwitch

aaa

reorder

aaa

---

Method clone()

The objects of this class are cloneable with this method.

Usage

PCPMr$clone(deep = FALSE)

Arguments

deep

Whether to make a deep clone.

Examples

# example code

# \donttest{
dat <- sim_J18K3phi3
#> Error: object 'sim_J18K3phi3' not found
J<-nrow(dat$lam);K<-ncol(dat$lam)
#> Error: object 'dat' not found
Q <- matrix(-1, nrow = J, ncol = K)
#> Error: object 'J' not found
Q[dat$lam!=0]=-2
#> Error: object 'Q' not found
Y<-dat$dat
#> Error: object 'dat' not found
a<-PCPMr$new()
a$pcfa(Y, Q, LD=F,iter=1000, burn=0,std=T, cor_fac=1, update = 1000)
#> Error: object 'Y' not found
out<-a$pcpmsummary(sig=F)

dat <- sim_J18K3P9
#> Error: object 'sim_J18K3P9' not found
J<-nrow(dat$lam);K<-ncol(dat$lam); P<-ncol(dat$mb)
#> Error: object 'dat' not found
Q <- matrix(-1, nrow = J, ncol = K)
#> Error: object 'J' not found
Q[dat$lam!=0]=-2
#> Error: object 'Q' not found
Qb <-matrix(-1,nrow = K, ncol = P)
#> Error: object 'K' not found
Y<-dat$dat[,1:J]
#> Error: object 'dat' not found
X<-dat$dat[,(J+1):(J+P)]
#> Error: object 'dat' not found
a<-PCPMr$new()
a$pclvm(Y, X, Q, Qb, LD=F, DIF=F,iter=1000, burn=0,std=T, cor_fac=1, update = 1000)
#> Error: attempt to apply non-function
# }