DPPCA: Dynamic Probabilistic Principal Components model (DPPCA)

Description Usage Arguments

View source: R/DPPCA.R

Description

DPCCA is a model for analysing longitudinal metabolomic data.

Usage

1
2
DPPCA(q, chain_output, prior_params, burn_in, thin, data_time,
  post_burn_in = 10, post_thin = 5)

Arguments

q

Number of components.

chain_output

Desired size of ouput chain.

prior_params

Named list containing prior parameter specification.
The parameters to be supplied are:
α, β: prior of v2, a IG(α/2, β/2) distribution.
μ_ν, σ^2_{ν}: prior of ν, a N(μ_ν,σ^2_{ν}) distribution.
μ_φ, σ^2_φ: prior of φ, a N_[-1,1](μ_φ,σ^2_φ) distribution.
α_V, β_V: prior of V, a IG((α_V)/2, (β_V)/2) distribution.
σ^2_μ: prior of μ, a N(0,σ^2_μ) distribution.
μ_Φ, σ^2_{Φ}: prior of Φ, a N_[-1,1](μ_Φ,σ^2_Φ) distribution.
Ω_m: prior covariance matrix of W_m, a MVN(0, Ω_m) distribution.

burn_in

length of burn-in period.

thin

thinning to be performed on chain.

data_time

List of M matrixes or data frames containing the observed data for each time point.
The data frame should contain observations in rows and spectral bins in columns.

post_burn_in

further burn-in applied to chains of loadings and scores to calculate posteriors

post_thin

further thinning applied to chains of loadings and scores to calculate posteriors


GweeXianYao/metaboliteR documentation built on Jan. 21, 2020, 7:18 a.m.