Package 'gomms'

GLM-Based Ordination Method

Description

preliminary analysis of similarity between samples in a low (2 or 3) dimensional display.

Author(s)

Michael B. Sohn

Maintainer: Michael B. Sohn <[email protected]>

References

Sohn, M.B. and Li, H. (2017). A GLM-Based Latent Variable Ordination Method for Microbiome Samples (Submitted).

Examples

## Not run: 
# load test data

data(gomms_test_data);

# estimate factor scores
cdat <- as.matrix(gomms_test_data[,-ncol(gomms_test_data)]);
rslt <- gomms(cdat);

# plot estimated factor scores
y <- as.matrix(gomms_test_data$group);
gomms.plot(rslt, y);

## End(Not run)

---

GLM-Based Ordination Method for Microbiome Samples

Description

estimate factor loadings and scores.

Usage

gomms(X, n.factors = 2, min.prop.nonzeros = 0.05, show.max.delta = FALSE)

Arguments

X	raw count data.
n.factors	number of factors. Default value is 2.
min.prop.nonzeros	minimum proportion of nonzeros allowed in analysis.
show.max.delta	display the maximum different between jth and (j+1)th iterations.

Value

estimated factor scores.

Author(s)

Michael B. Sohn

Maintainer: Michael B. Sohn <[email protected]>

References

Sohn, M.B. and Li, H. (2017). A GLM-Based Latent Variable Ordination Method for Microbiome Samples (Submitted).

---

Test Data

Description

70 samples and 83 features. The last column contains the population identification for each sample.

Usage

data(gomms_test_data)

---

Plot Factor Loadings

Description

plot estimated factor loadings for each sample.

Usage

gomms.plot(X, Y, col.markers = NULL, pch.markers = NULL, ...)

Arguments

X	two dimnsional matrix of factor scores.
Y	one or two dimensional matrix of classification.
col.markers	user specified colors for classification.
pch.markers	user specified plot symbols for classification.
...	optional graphical parameters to be passed.

---

Probability of a Zero from a Zero State

Description

estimate the probability of a zero from a zero state.

Usage

Qqpois(cdat, eta.hat, mu.hat, dispersion)

Arguments

cdat	count Data.
eta.hat	estimated proportion of zeros from a zero state.
mu.hat	estimated mean count.
dispersion	estimated values for dispersion.

Value

estimated probability of a zero from a zero state.

---