adpc {adproclus}R Documentation

Constructor for a (low dimensional) ADPROCLUS solution object

Description

Yields an object of class adpc, which can be printed, plotted and summarized by the corresponding methods. Mandatory input are the membership matrix AA and the profile matrix PP (where the number of columns from AA corresponds to the number of rows in PP), if the object is to represent a full dimensional ADPROCLUS model. For a low dimensional ADPROCLUS model, the matrices CC and BB have to be provided and PP can be inferred from those. All other inputs are optional but may be included so that the output from the summary(), print(), plot() is complete. For further details on the (low dimensional) ADPROCLUS model and what every element of the objects means see adproclus and adproclus_low_dim.

Usage

adpc(
  A,
  P,
  sse = NULL,
  totvar = NULL,
  explvar = NULL,
  iterations = NULL,
  timer = NULL,
  timer_one_run = NULL,
  initial_start = NULL,
  C = NULL,
  B = NULL,
  runs = NULL,
  parameters = NULL
)

Arguments

A

Membership matrix A.

P

Profile matrix P.

sse

Sum of Squared Error.

totvar

Total variance.

explvar

Explained variance.

iterations

Number of iterations.

timer

Time needed to run the complete algorithm.

timer_one_run

Time to complete this single algorithm start.

initial_start

List containing type of start and start_allocation matrix.

C

Low dimensional profiles matrix C.

B

Matrix of base vectors connecting low dimensional components with original variables B.

runs

List of suboptimal models.

parameters

List of algorithm parameters.

Value

Object of class adpc.

Examples

# Create the information needed for a minimal object of class adpc
x <- stackloss
result <- adproclus(x, 3)
A <- result$A
P <- result$P

# Use constructor to obtain object of class adpc
result_object <- adpc(A, P)


[Package adproclus version 1.0.2 Index]