hdpSampleMulti-class: hdpSampleMulti class for multiple independent hdpSampleChain...

Description Usage Arguments Methods (by generic) Slots

Description

hdpSampleMulti class for multiple independent hdpSampleChain objects for the same HDP

Usage

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
## S4 method for signature 'hdpSampleMulti'
as.list(x, ...)

## S4 method for signature 'hdpSampleMulti'
chains(x, ...)

## S4 method for signature 'hdpSampleMulti'
numcomp(x)

## S4 method for signature 'hdpSampleMulti'
prop.ex(x)

## S4 method for signature 'hdpSampleMulti'
comp_cos_merge(x)

## S4 method for signature 'hdpSampleMulti'
comp_categ_counts(x)

## S4 method for signature 'hdpSampleMulti'
comp_dp_counts(x)

## S4 method for signature 'hdpSampleMulti'
comp_categ_distn(x)

## S4 method for signature 'hdpSampleMulti'
comp_dp_distn(x)

Arguments

x

Object of class hdpSampleMulti

...

unused

Methods (by generic)

Slots

chains

List of hdpSampleChain objects storing multiple independent runs of the posterior sampling chain for the same data and HDP struct

numcomp

Number of global components extracted by hdp_extract_components (not including component 0)

prop.ex

(Average) proportion of dataset explained by the extracted components

comp_cos_merge

cos.merge setting used by hdp_extract_components

comp_categ_counts

List of matrices (one for each component) counting the sample-category data assignment across all DP nodes. Number of rows is the number of posterior samples, and number of columns is the number of data categories.

comp_dp_counts

List of matrices (one for each DP) counting sample-component assignment (aggregating across data categories). Number of rows is the number of posterior samples, and number of columns is the number of components.

comp_categ_distn

List with elements "mean" and "cred.int", containing matrices with the mean (and lower/upper 95% credibility interval) distribution over data categories for each component. Number of rows is the number of components, and number of columns is the number of data categories. Rows sum to 1.

comp_dp_distn

List with elements "mean" and "cred.int", containing matrices with the mean (and lower/upper 95% credibility interval) distribution over components for each DP. Number of rows is the number of DPs, and number of columns is the number of components. Rows sum to 1.


nicolaroberts/hdp documentation built on May 23, 2019, 5:09 p.m.