library(knitr) library(st4gi) #library(rmdformats) traits <- params$trait treat <- params$treat rep <- params$rep data <- params$data maxp <- params$maxp meta <- params$meta host <- params$host geno <- treat
# This is an automatedly created report. # See more details in section on materials.
phs_lbl = "Advanced Trial" ttl <- stringr::str_sub(meta$title, 1, 2) if (stringr::str_detect(ttl, "PT")) {phs_lbl = "Preliminary Trial"} if (stringr::str_detect(ttl, "OT")) {phs_lbl = "Observation Trial"} brp_lbl = "Yield Breeding Program"
This trial has the identifier r meta$title
. It was conducted under the supervision of r meta$contact
as a r phs_lbl
as part of a r brp_lbl
in r meta$site
, r meta$country
in r meta$year
. A total of r length(unique(data[, treat]))
clones (including reference clones) were evaluated for r length(params$trait)
traits.
There is data from r length(unique(data[, treat]))
treatments, evaluated using a randomize complete block design with r unique(data[, rep])
blocks. The statistical model is
$$
y_{ij} = \mu + \tau_i + \beta_j + \epsilon_{ij}
$$
where
In this model we assume that the errors are independent and have a normal distribution with common variance, that is, $\epsilon_{ij} \sim N(0,\sigma_{\epsilon}^2)$.
The following traits are analyzed: r paste(params$trait, collapse = ", ")
.
gid = unique(data$germplasmDbId) gnm = unique(data$germplasmName) path = "/stock/" #TODO change for genotypes out = paste0("<a href='http://",host, path, gid,"/view' target='_blank'>",gnm,"</a>") txt = paste0("") # TODO make trait choosable out = paste( out, collapse = ", ") gidOut = paste(txt, out)
The following germplasm was analyzed: r paste(gidOut)
.
s <- sessionInfo()
This report was created using r s$R.version
on a r s$platform
running r s$running
in r s$locacel
. The following base packages were loaded:
r paste(unlist(s$basePkgs), collapse = ", ")
and the following additional packages:
r paste(names(s$otherPkgs), collapse = ", ")
.
data = data[, c(treat, "REP", traits)] # data[, treat] <- as.factor(data[, treat]) # exclude the response variable and empty variable for RF imputation datas <- names(data)[!names(data) %in% c(treat, "PED1")] # TODO replace "PED1" by a search #datas <- names(data)[!names(data) %in% c(treat)] x <- data[, datas] # x <- data[, trait] for(i in 1:ncol(x)){ x[, i] <- as.numeric(x[, i]) } # y <- data[, treat] # determine which traits are having more than 10% missing values mval = 0.1 n = ncol(data) m = nrow(data) off = round(mval * m, 0) mval = logical(n) mvan = numeric(n) # Remove all complete NA first # for(i in 1:n) { # #mvan[i] = nrow(data[is.na(data[, i]), ]) # mval[i] = nrow(data[is.na(data[, i]), ]) == nrow(data) # } for(i in 1:n) { mvan[i] = nrow(data[is.na(data[, i]), ]) mval[i] = nrow(data[is.na(data[, i]), ]) / m * 100 <= off } mvnm = names(data)[!mval] dat = data if (any(is.na(x))){ #capture.output({ dat <- randomForest::rfImpute(x = x, y = y ) names(dat)[1] <- treat dat = dat[, mval] } #names(data)[1] <- "REP" # xnm = names(dat) trts = names(dat)[-c(1,2)] dat[, treat] <- as.character(dat[, treat]) #lc <- st4gi::checkdata01(trts, treat, rep, dat) tbl = dat
The following traits were not analyzed since they had too many missing values (>= 10%): r paste(mvnm)
. For the remaining traits missing values were imputed using all available information.
Valid traits: r paste(names(dat)[-c(1,2)])
.
out <- NULL for (k in 1:length(trts)) { lc <- st4gi::checkdata01(trts[k], treat, rep, tbl) out <- c(out, knit_expand('child_rcbd2_old.Rmd')) }
r paste(knit(text = out), collapse = '\n')
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.