cpa_getOTW_2: Obtain the Optimal Time Window (OTW)

In ICRISAT-GEMS/SpaTemHTP: Processing and utilization of HTP data

Description

Perform Change Point Analysis (CPA) on genotype adjusted means or G-BLUEs time series of a trait

Usage

cpa_getOTW_2(data, h2)

Arguments

data	Numeric Matrix of dimension (N_genotype x N_days) containing the G-BLUEs of a particular trait. The data colnmanes must be in Date compatible format 'dd-mm-yyyy'.
h2	Numeric vector of dimension (N_days) containing the heritability estimates on each day for the trait

Details

Entire temporal data set is partitioned into windows based on the differences in the distribution of trait heritability and phenotypic separability using a change point analysis procedure defined by Matteson et al. (2014).

Value

List with the following items:

change_points	Numeric vector indicating the location of the change points on the time series.
change_points_dates	Character vector dates of the time series change points.
OTW_data	List of Numeric matrices representing the genotype trait values within the OTW.
OTW_data_opt	Numeric matrix with the genotype trait values within the OTW with the highest heritability.

Author(s)

Soumyashree Kar, Vincent Garin

References

Matteson, D.S. and James, N.A. (2014). A nonparametric approach for multiple change point analysis of multivariate data. Journal of the American Statistical Association, 109(505), pp.334-345.

Examples

data(SG_PH_data)
SG_PH_data$col_f <- factor(SG_PH_data$col)
SG_PH_data$row_f <- factor(SG_PH_data$row)

SG_PH_data$rep <- factor(SG_PH_data$rep)
SG_PH_data$block <- factor(SG_PH_data$block)

exp_des_data = SG_PH_data[, c("row", "col", "row_f", "col_f","genotype",
                              "rep", "block")]

## Not run: 

op <- SpaTemHTP_proc(exp_des_data, pheno_data = SG_PH_data[, 6:28],
                     out_det = TRUE, miss_imp = TRUE, sp_adj = TRUE,
                     random = ~ rep +  rep:block + row_f + col_f,
                     h2_comp = TRUE, plot = TRUE)
                     
data <- op$G_BLUES

# make sure data colnmanes are dd-mm-yyyy Date format compatible
dates <- substr(colnames(data), 2, nchar(colnames(data)))
dates <- str_replace_all(string = dates, pattern = '\\.', replacement = '-')
colnames(data) <- dates

h2 <- op$h2

OTW <- cpa_getOTW_2(data = data, h2 = h2)


## End(Not run)

ICRISAT-GEMS/SpaTemHTP documentation built on March 9, 2021, 12:12 a.m.