In NCI-CGR/cgrtelomeres: Automated Processing of Telomere qPCR Data
Set global ggplot2 theme
my.theme <- ggplot2::theme_light() +
ggplot2::theme(
plot.title = ggplot2::element_text(size = 14, hjust = 0.5),
axis.title = ggplot2::element_text(size = 12),
axis.text = ggplot2::element_text(size = 10),
legend.title = ggplot2::element_text(size = 12),
legend.text = ggplot2::element_text(size = 10),
strip.background = ggplot2::element_blank(),
strip.text = ggplot2::element_text(size = 12, colour = "black")
)
Make sure required output directory structure exists
## if output directory path does not exist, create it
create.output.directories(output.path)
Detect input files based on fixed input directory structure
## aggregate pairs of filenames from input.path/Data/Exports
input.files <- find.input.files(paste(input.path, "Data",
"Exports",
sep = "/"
))
Load data from detected input files
## load data from acquired pairs of files
input.data <- lapply(input.files, function(i) {
read.export.datum(i,
source.search.path = paste(input.path,
"Data",
"Analysis",
sep = "/"
),
subject.list.from.input.path = subject.list.from.input.path,
plate.content.report = plate.content.report,
plate.list = plate.list
)
})
Run primary analysis generation on loaded data
## run primary analysis steps for Data/Analysis
primary.analysis <- lapply(input.data, function(i) {
create.analysis(i,
plate.content.report = plate.content.report,
plate.list = plate.list,
infer.384.locations = infer.384.locations
)
})
Report primary analysis results to file
## report the primary analysis results to the output/Data/Analysis
lapply(primary.analysis, function(i) {
report.primary.analysis(i, output.path, project.id)
})
Report final summary table
## report the final summary table of combined data across primary analyses
res <- rbind()
for (i in 1:length(primary.analysis)) {
res <- rbind(
res,
format.final.analysis(
primary.analysis[[i]],
project.id
)
)
}
is.techrep <- duplicated(res[, "Sample ID"]) |
duplicated(res[, "Sample ID"], fromLast = TRUE)
is.ntc <- res[, "Sample Type"] == "Negative Control"
is.control <- res[, "Sample Type"] == "Internal Control"
res[is.techrep &
!is.ntc &
!is.control, "Sample Type"] <-
"Technical Replicate"
res <- res[!is.na(res[, "Vial ID"]) &
!is.na(res[, "Sample ID"]), ]
write.table(res,
final.tsv,
row.names = FALSE,
col.names = TRUE,
quote = FALSE,
sep = "\t"
)
Compute ICC
## extract subject dups only
res.dups <- res[res[, "Vial ID"] != "NA07057" &
res[, "Sample ID"] != "NTC", ]
completion.rate <- length(which(!is.na(res.dups[, "Standardized T/S Ratio"])))
completion.rate <- completion.rate / nrow(res.dups) * 100
telo.cv <- mean(res.dups[, "Telo %CV"], na.rm = TRUE)
control.cv <- mean(res.dups[, "36B4 %CV"], na.rm = TRUE)
unique.study.samples <- length(unique(res.dups[, "Sample ID"]))
res.dups <- res.dups[duplicated(res.dups[, "Sample ID"]) |
duplicated(res.dups[, "Sample ID"], fromLast = TRUE), ]
n.tech.reps <- length(unique(res.dups[, "Sample ID"]))
## compute ICC to match existing method
ICC.est <- ICC::ICCest(
factor(res.dups[, "Sample ID"]),
res.dups[, "Standardized T/S Ratio"]
)
Report first summary text
Samples
A total of r unique.study.samples unique study samples were run
in this project. CGR Internal Controls (NA07057) were added to five
of those wells and water (NTC) was added to one of those wells.
A total of r nrow(res) sample instances in the project were tested
across r length(unique(res[, "Intermediate Source Plate ID"]))
sets of plates.
- Overall completion rate of study samples:
r round(completion.rate, 2)%
- Average %CV for Telo assay triplicates:
r round(telo.cv, 3)%
- Average %CV for 36B4 assay triplicates:
r round(control.cv, 3)%
- Technical replicates in project (n=
r n.tech.reps),
r round(n.tech.reps/unique.study.samples*100,1)%
Report per-plate exponential fit
Plate Standard Curves
fit.group <- c()
fit.intercept.telo <- c()
fit.slope.telo <- c()
fit.r2.telo <- c()
fit.intercept.control <- c()
fit.slope.control <- c()
fit.r2.control <- c()
fixed.concs <- log2(c(4, 1.6, 0.64, 0.256, 0.1024, 0.041))
for (i in 1:length(primary.analysis)) {
pa <- primary.analysis[[i]]
telo <- log2(pa@Avg.ExperimentalCt[1:6])
cont <- log2(pa@Avg.ControlCt[1:6])
int.telo <- summary(pa@Model.Experiment)$coeff[1, 1]
slope.telo <- summary(pa@Model.Experiment)$coeff[2, 1]
r2.telo <- summary(pa@Model.Experiment)$r.squared
int.cont <- summary(pa@Model.Control)$coeff[1, 1]
slope.cont <- summary(pa@Model.Control)$coeff[2, 1]
r2.cont <- summary(pa@Model.Control)$r.squared
telo.data <- data.frame(
x = fixed.concs,
y = telo
)
plot.telo <- ggplot2::ggplot(ggplot2::aes(
x = .data$x,
y = .data$y
),
data = telo.data
)
plot.x <- fixed.concs[1]
plot.xend <- fixed.concs[length(fixed.concs)]
plot.y <- int.telo + slope.telo * fixed.concs[1]
plot.yend <- int.telo + slope.telo *
fixed.concs[length(fixed.concs)]
plot.telo <- plot.telo + my.theme
plot.telo <- plot.telo + ggplot2::geom_point(colour = "blue")
plot.telo <- plot.telo + ggplot2::geom_segment(
x = plot.x,
xend = plot.xend,
y = plot.y,
yend = plot.yend
)
plot.telo <- plot.telo + ggplot2::scale_y_continuous(limits = c(0, 5))
plot.telo <- plot.telo + ggplot2::xlab(expression("log"[2] *
"(standard concentrations)"))
plot.telo <- plot.telo + ggplot2::ylab(expression("log"[2] *
"(telomere observations)"))
plot.telo <- plot.telo + ggplot2::ggtitle(paste(project.id, "_",
pa@Analysis.Code,
" Telomere Curve",
sep = ""
))
print(plot.telo)
cont.data <- data.frame(
x = fixed.concs,
y = cont
)
plot.cont <- ggplot2::ggplot(ggplot2::aes(
x = .data$x,
y = .data$y
),
data = cont.data
)
plot.x <- fixed.concs[1]
plot.xend <- fixed.concs[length(fixed.concs)]
plot.y <- int.cont + slope.cont * fixed.concs[1]
plot.yend <- int.cont + slope.cont *
fixed.concs[length(fixed.concs)]
plot.cont <- plot.cont + my.theme
plot.cont <- plot.cont + ggplot2::geom_point(colour = "red")
plot.cont <- plot.cont + ggplot2::geom_segment(
x = plot.x,
xend = plot.xend,
y = plot.y,
yend = plot.yend
)
plot.cont <- plot.cont + ggplot2::scale_y_continuous(limits = c(0, 5))
plot.cont <- plot.cont + ggplot2::xlab(expression("log"[2] *
"(standard concentrations)"))
plot.cont <- plot.cont + ggplot2::ylab(expression("log"[2] *
"(36B4 observations)"))
plot.cont <- plot.cont + ggplot2::ggtitle(paste(project.id, "_",
pa@Analysis.Code,
" 36B4 Curve",
sep = ""
))
print(plot.cont)
fit.group <- c(
fit.group,
paste(project.id, "_", pa@Analysis.Code, sep = "")
)
fit.intercept.telo <- c(
fit.intercept.telo,
int.telo
)
fit.slope.telo <- c(
fit.slope.telo,
slope.telo
)
fit.r2.telo <- c(
fit.r2.telo,
r2.telo
)
fit.intercept.control <- c(
fit.intercept.control,
int.cont
)
fit.slope.control <- c(
fit.slope.control,
slope.cont
)
fit.r2.control <- c(
fit.r2.control,
r2.cont
)
}
fit.report <- data.frame(
Group = fit.group,
Telomere.Intercept = fit.intercept.telo,
Telomere.Slope = fit.slope.telo,
Telomere.r2 = fit.r2.telo,
Control.Intercept = fit.intercept.control,
Control.Slope = fit.slope.control,
Control.r2 = fit.r2.control
)
print(fit.report)
Compute information about internal controls
res.int <- res[res[, "Vial ID"] == "NA07057", ]
mean.cv.data <- res.int[, "Raw T/S Ratio"]
mean.cv <- 100 * sd(mean.cv.data, na.rm = TRUE) /
mean(mean.cv.data, na.rm = TRUE)
box.data <- data.frame(
x = factor(rep(res.int[, "Intermediate Source Plate ID"], 2)),
y = c(
res.int[, "Raw T/S Ratio"],
res.int[, "Standardized T/S Ratio"]
),
outtype = rep(c(
"Raw T/S Ratio",
"Standardized T/S Ratio"
),
each = nrow(res.int)
)
)
cv.box <- ggplot2::ggplot(ggplot2::aes(
x = .data$x,
y = .data$y,
outtype = .data$outtype
),
data = box.data
)
cv.box <- cv.box + my.theme
cv.box <- cv.box + ggplot2::geom_boxplot(ggplot2::aes(
x = .data$x,
y = .data$y
))
cv.box <- cv.box + ggplot2::geom_jitter(
shape = 16,
position = ggplot2::position_jitter(0.2)
)
cv.box <- cv.box + ggplot2::scale_y_continuous(
limits = c(0, 3),
breaks = seq(0, 3, 0.5)
)
cv.box <- cv.box + ggplot2::theme(
axis.text.x =
ggplot2::element_text(
size = 10,
angle = 90
)
)
cv.box <- cv.box + ggplot2::xlab("Intermediate Source Plate ID")
cv.box <- cv.box + ggplot2::ylab("")
cv.box <- cv.box + ggplot2::facet_grid(.data$outtype ~ .)
Report on internal controls
Internal Controls
The CGR Internal Control, which is used to standardize the raw T/S ratios for all samples
within each plate of the project, had an overall CV of r round(mean.cv, 3)%.
Render the CV boxplot
print(cv.box)
Compute information on technical replicates
techrep.data <- data.frame(
y = res.dups[, "Standardized T/S Ratio"],
x = res.dups[, "Sample ID"]
)
## assign each sample its average value for plotting
tech.samples <- unique(techrep.data$x)
tech.values <- c()
for (id in tech.samples) {
tech.values <- c(
tech.values,
mean(techrep.data[techrep.data$x == id, "y"],
na.rm = TRUE
)
)
}
tech.samples <- tech.samples[order(tech.values)]
techrep.data$x <- factor(techrep.data$x, levels = tech.samples)
techrep.plot <- ggplot2::ggplot(ggplot2::aes(
x = .data$x,
y = .data$y
),
data = techrep.data
)
techrep.plot <- techrep.plot + my.theme
techrep.plot <- techrep.plot + ggplot2::theme(axis.text.x = ggplot2::element_blank())
techrep.plot <- techrep.plot + ggplot2::geom_point()
techrep.plot <- techrep.plot + ggplot2::xlab("")
techrep.plot <- techrep.plot + ggplot2::ylab("Standardized T/S Ratio")
Report on Technical Replicates
The distribution of standardized T/S ratios from the technical replicates
between plates can be seen below. The intraclass correlation coefficient (ICC)
and its 95% confidence interval were r round(ICC.est$ICC, 3)
(r round(ICC.est$LowerCI, 3),r round(ICC.est$UpperCI, 3)).
Render the technical replicate scatterplot
print(techrep.plot)
Report technical replicate data for easier lookup
first.techrep <- techrep.data[duplicated(techrep.data$x), ]
first.techrep$x <- as.vector(first.techrep$x, mode = "character")
first.techrep <- first.techrep[order(first.techrep$x), ]
second.techrep <- techrep.data[duplicated(techrep.data$x, fromLast = TRUE), ]
second.techrep$x <- as.vector(second.techrep$x, mode = "character")
second.techrep <- second.techrep[order(second.techrep$x), ]
combined.techrep <- data.frame(
Sample.ID = first.techrep$x,
Smaller.Value = first.techrep$y,
Larger.Value = second.techrep$y
)
combined.na <- combined.techrep[is.na(combined.techrep$Smaller.Value) |
is.na(combined.techrep$Larger.Value), ]
combined.techrep <- combined.techrep[!is.na(combined.techrep$Smaller.Value) &
!is.na(combined.techrep$Larger.Value), ]
problem.ind <- combined.techrep$Smaller.Value > combined.techrep$Larger.Value
tmp <- combined.techrep$Smaller.Value[problem.ind]
combined.techrep$Smaller.Value[problem.ind] <-
combined.techrep$Larger.Value[problem.ind]
combined.techrep$Larger.Value[problem.ind] <- tmp
print(rbind(combined.techrep, combined.na))
NCI-CGR/cgrtelomeres documentation built on Feb. 11, 2021, 12:12 p.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Set global ggplot2 theme
my.theme <- ggplot2::theme_light() + ggplot2::theme( plot.title = ggplot2::element_text(size = 14, hjust = 0.5), axis.title = ggplot2::element_text(size = 12), axis.text = ggplot2::element_text(size = 10), legend.title = ggplot2::element_text(size = 12), legend.text = ggplot2::element_text(size = 10), strip.background = ggplot2::element_blank(), strip.text = ggplot2::element_text(size = 12, colour = "black") )
Make sure required output directory structure exists
## if output directory path does not exist, create it create.output.directories(output.path)
Detect input files based on fixed input directory structure
## aggregate pairs of filenames from input.path/Data/Exports input.files <- find.input.files(paste(input.path, "Data", "Exports", sep = "/" ))
Load data from detected input files
## load data from acquired pairs of files input.data <- lapply(input.files, function(i) { read.export.datum(i, source.search.path = paste(input.path, "Data", "Analysis", sep = "/" ), subject.list.from.input.path = subject.list.from.input.path, plate.content.report = plate.content.report, plate.list = plate.list ) })
Run primary analysis generation on loaded data
## run primary analysis steps for Data/Analysis primary.analysis <- lapply(input.data, function(i) { create.analysis(i, plate.content.report = plate.content.report, plate.list = plate.list, infer.384.locations = infer.384.locations ) })
Report primary analysis results to file
## report the primary analysis results to the output/Data/Analysis lapply(primary.analysis, function(i) { report.primary.analysis(i, output.path, project.id) })
Report final summary table
## report the final summary table of combined data across primary analyses res <- rbind() for (i in 1:length(primary.analysis)) { res <- rbind( res, format.final.analysis( primary.analysis[[i]], project.id ) ) } is.techrep <- duplicated(res[, "Sample ID"]) | duplicated(res[, "Sample ID"], fromLast = TRUE) is.ntc <- res[, "Sample Type"] == "Negative Control" is.control <- res[, "Sample Type"] == "Internal Control" res[is.techrep & !is.ntc & !is.control, "Sample Type"] <- "Technical Replicate" res <- res[!is.na(res[, "Vial ID"]) & !is.na(res[, "Sample ID"]), ] write.table(res, final.tsv, row.names = FALSE, col.names = TRUE, quote = FALSE, sep = "\t" )
Compute ICC
## extract subject dups only res.dups <- res[res[, "Vial ID"] != "NA07057" & res[, "Sample ID"] != "NTC", ] completion.rate <- length(which(!is.na(res.dups[, "Standardized T/S Ratio"]))) completion.rate <- completion.rate / nrow(res.dups) * 100 telo.cv <- mean(res.dups[, "Telo %CV"], na.rm = TRUE) control.cv <- mean(res.dups[, "36B4 %CV"], na.rm = TRUE) unique.study.samples <- length(unique(res.dups[, "Sample ID"])) res.dups <- res.dups[duplicated(res.dups[, "Sample ID"]) | duplicated(res.dups[, "Sample ID"], fromLast = TRUE), ] n.tech.reps <- length(unique(res.dups[, "Sample ID"])) ## compute ICC to match existing method ICC.est <- ICC::ICCest( factor(res.dups[, "Sample ID"]), res.dups[, "Standardized T/S Ratio"] )
Report first summary text
Samples
A total of r unique.study.samples unique study samples were run
in this project. CGR Internal Controls (NA07057) were added to five
of those wells and water (NTC) was added to one of those wells.
A total of r nrow(res) sample instances in the project were tested
across r length(unique(res[, "Intermediate Source Plate ID"]))
sets of plates.
- Overall completion rate of study samples:
r round(completion.rate, 2)% - Average %CV for Telo assay triplicates:
r round(telo.cv, 3)% - Average %CV for 36B4 assay triplicates:
r round(control.cv, 3)% - Technical replicates in project (n=
r n.tech.reps),r round(n.tech.reps/unique.study.samples*100,1)%
Report per-plate exponential fit
Plate Standard Curves
fit.group <- c() fit.intercept.telo <- c() fit.slope.telo <- c() fit.r2.telo <- c() fit.intercept.control <- c() fit.slope.control <- c() fit.r2.control <- c() fixed.concs <- log2(c(4, 1.6, 0.64, 0.256, 0.1024, 0.041)) for (i in 1:length(primary.analysis)) { pa <- primary.analysis[[i]] telo <- log2(pa@Avg.ExperimentalCt[1:6]) cont <- log2(pa@Avg.ControlCt[1:6]) int.telo <- summary(pa@Model.Experiment)$coeff[1, 1] slope.telo <- summary(pa@Model.Experiment)$coeff[2, 1] r2.telo <- summary(pa@Model.Experiment)$r.squared int.cont <- summary(pa@Model.Control)$coeff[1, 1] slope.cont <- summary(pa@Model.Control)$coeff[2, 1] r2.cont <- summary(pa@Model.Control)$r.squared telo.data <- data.frame( x = fixed.concs, y = telo ) plot.telo <- ggplot2::ggplot(ggplot2::aes( x = .data$x, y = .data$y ), data = telo.data ) plot.x <- fixed.concs[1] plot.xend <- fixed.concs[length(fixed.concs)] plot.y <- int.telo + slope.telo * fixed.concs[1] plot.yend <- int.telo + slope.telo * fixed.concs[length(fixed.concs)] plot.telo <- plot.telo + my.theme plot.telo <- plot.telo + ggplot2::geom_point(colour = "blue") plot.telo <- plot.telo + ggplot2::geom_segment( x = plot.x, xend = plot.xend, y = plot.y, yend = plot.yend ) plot.telo <- plot.telo + ggplot2::scale_y_continuous(limits = c(0, 5)) plot.telo <- plot.telo + ggplot2::xlab(expression("log"[2] * "(standard concentrations)")) plot.telo <- plot.telo + ggplot2::ylab(expression("log"[2] * "(telomere observations)")) plot.telo <- plot.telo + ggplot2::ggtitle(paste(project.id, "_", pa@Analysis.Code, " Telomere Curve", sep = "" )) print(plot.telo) cont.data <- data.frame( x = fixed.concs, y = cont ) plot.cont <- ggplot2::ggplot(ggplot2::aes( x = .data$x, y = .data$y ), data = cont.data ) plot.x <- fixed.concs[1] plot.xend <- fixed.concs[length(fixed.concs)] plot.y <- int.cont + slope.cont * fixed.concs[1] plot.yend <- int.cont + slope.cont * fixed.concs[length(fixed.concs)] plot.cont <- plot.cont + my.theme plot.cont <- plot.cont + ggplot2::geom_point(colour = "red") plot.cont <- plot.cont + ggplot2::geom_segment( x = plot.x, xend = plot.xend, y = plot.y, yend = plot.yend ) plot.cont <- plot.cont + ggplot2::scale_y_continuous(limits = c(0, 5)) plot.cont <- plot.cont + ggplot2::xlab(expression("log"[2] * "(standard concentrations)")) plot.cont <- plot.cont + ggplot2::ylab(expression("log"[2] * "(36B4 observations)")) plot.cont <- plot.cont + ggplot2::ggtitle(paste(project.id, "_", pa@Analysis.Code, " 36B4 Curve", sep = "" )) print(plot.cont) fit.group <- c( fit.group, paste(project.id, "_", pa@Analysis.Code, sep = "") ) fit.intercept.telo <- c( fit.intercept.telo, int.telo ) fit.slope.telo <- c( fit.slope.telo, slope.telo ) fit.r2.telo <- c( fit.r2.telo, r2.telo ) fit.intercept.control <- c( fit.intercept.control, int.cont ) fit.slope.control <- c( fit.slope.control, slope.cont ) fit.r2.control <- c( fit.r2.control, r2.cont ) } fit.report <- data.frame( Group = fit.group, Telomere.Intercept = fit.intercept.telo, Telomere.Slope = fit.slope.telo, Telomere.r2 = fit.r2.telo, Control.Intercept = fit.intercept.control, Control.Slope = fit.slope.control, Control.r2 = fit.r2.control ) print(fit.report)
Compute information about internal controls
res.int <- res[res[, "Vial ID"] == "NA07057", ] mean.cv.data <- res.int[, "Raw T/S Ratio"] mean.cv <- 100 * sd(mean.cv.data, na.rm = TRUE) / mean(mean.cv.data, na.rm = TRUE) box.data <- data.frame( x = factor(rep(res.int[, "Intermediate Source Plate ID"], 2)), y = c( res.int[, "Raw T/S Ratio"], res.int[, "Standardized T/S Ratio"] ), outtype = rep(c( "Raw T/S Ratio", "Standardized T/S Ratio" ), each = nrow(res.int) ) ) cv.box <- ggplot2::ggplot(ggplot2::aes( x = .data$x, y = .data$y, outtype = .data$outtype ), data = box.data ) cv.box <- cv.box + my.theme cv.box <- cv.box + ggplot2::geom_boxplot(ggplot2::aes( x = .data$x, y = .data$y )) cv.box <- cv.box + ggplot2::geom_jitter( shape = 16, position = ggplot2::position_jitter(0.2) ) cv.box <- cv.box + ggplot2::scale_y_continuous( limits = c(0, 3), breaks = seq(0, 3, 0.5) ) cv.box <- cv.box + ggplot2::theme( axis.text.x = ggplot2::element_text( size = 10, angle = 90 ) ) cv.box <- cv.box + ggplot2::xlab("Intermediate Source Plate ID") cv.box <- cv.box + ggplot2::ylab("") cv.box <- cv.box + ggplot2::facet_grid(.data$outtype ~ .)
Report on internal controls
Internal Controls
The CGR Internal Control, which is used to standardize the raw T/S ratios for all samples
within each plate of the project, had an overall CV of r round(mean.cv, 3)%.
Render the CV boxplot
print(cv.box)
Compute information on technical replicates
techrep.data <- data.frame( y = res.dups[, "Standardized T/S Ratio"], x = res.dups[, "Sample ID"] ) ## assign each sample its average value for plotting tech.samples <- unique(techrep.data$x) tech.values <- c() for (id in tech.samples) { tech.values <- c( tech.values, mean(techrep.data[techrep.data$x == id, "y"], na.rm = TRUE ) ) } tech.samples <- tech.samples[order(tech.values)] techrep.data$x <- factor(techrep.data$x, levels = tech.samples) techrep.plot <- ggplot2::ggplot(ggplot2::aes( x = .data$x, y = .data$y ), data = techrep.data ) techrep.plot <- techrep.plot + my.theme techrep.plot <- techrep.plot + ggplot2::theme(axis.text.x = ggplot2::element_blank()) techrep.plot <- techrep.plot + ggplot2::geom_point() techrep.plot <- techrep.plot + ggplot2::xlab("") techrep.plot <- techrep.plot + ggplot2::ylab("Standardized T/S Ratio")
Report on Technical Replicates
The distribution of standardized T/S ratios from the technical replicates
between plates can be seen below. The intraclass correlation coefficient (ICC)
and its 95% confidence interval were r round(ICC.est$ICC, 3)
(r round(ICC.est$LowerCI, 3),r round(ICC.est$UpperCI, 3)).
Render the technical replicate scatterplot
print(techrep.plot)
Report technical replicate data for easier lookup
first.techrep <- techrep.data[duplicated(techrep.data$x), ] first.techrep$x <- as.vector(first.techrep$x, mode = "character") first.techrep <- first.techrep[order(first.techrep$x), ] second.techrep <- techrep.data[duplicated(techrep.data$x, fromLast = TRUE), ] second.techrep$x <- as.vector(second.techrep$x, mode = "character") second.techrep <- second.techrep[order(second.techrep$x), ] combined.techrep <- data.frame( Sample.ID = first.techrep$x, Smaller.Value = first.techrep$y, Larger.Value = second.techrep$y ) combined.na <- combined.techrep[is.na(combined.techrep$Smaller.Value) | is.na(combined.techrep$Larger.Value), ] combined.techrep <- combined.techrep[!is.na(combined.techrep$Smaller.Value) & !is.na(combined.techrep$Larger.Value), ] problem.ind <- combined.techrep$Smaller.Value > combined.techrep$Larger.Value tmp <- combined.techrep$Smaller.Value[problem.ind] combined.techrep$Smaller.Value[problem.ind] <- combined.techrep$Larger.Value[problem.ind] combined.techrep$Larger.Value[problem.ind] <- tmp print(rbind(combined.techrep, combined.na))
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.