Related
My example data is as follows:
df <- data.frame(study = c("Hodaie","Kerrigan","Lee","Andrade","Lim"), SR = c(0.5460, 0.2270, 0.7540, 0.6420, 0.5000), SE = c(12.30, 15.70, 12.80, 13.80, 9.00), Patients = c(5, 5, 3, 6, 4))
I want to conduct the meta-analysis with SR(single group percentage), SE (standard error that I can compute based on sample size and percentage), and patients(sample size for each study), and I hope I could get the following forest plot (I found this example in an article, and they also have one group percentage data, but I can't find which R statement or argument they used):
Could anyone tell me which R statement or argument that I could use to conduct the meta-analysis and generate the forest plot above? Thank you!
I am sure there are plenty of ways to do this using packages but it can be accomplished in base R (and there are likely more elegant solutions using base R). The way I do it is to first build a blank plot much larger than the needed graphing portion, then overlay the relevant elements on it. I find one has more control over it this way. A basic example that could get you started is below. If you are new to R (based on your name NewRUser), I suggest running it line-by-line to see how it all works. Again, this is only one way and there are likely better approaches. Good luck!
Sample Data
#### Sample Data (modified from OP)
df <- data.frame(Study = c("Hodaie","Kerrigan","Lee","Andrade","Lim"),
SR = c(0.5460, 0.2270, 0.7540, 0.6420, 0.5000),
SE = c(12.30, 15.70, 12.80, 13.80, 9.00),
Patients = c(5, 5, 3, 6, 4),
ci_lo = c(30, -8.0, 50, 37, 32),
ci_hi = c(78, 53, 100, 91, 67))
### Set up plotting elements
n.studies <- nrow(df)
yy <- n.studies:1
seqx <- seq(-100, 100, 50)
## blank plot much larger than needed
plot(range(-550, 200), range(0, n.studies), type = 'n', axes = F, xlab = '', ylab = '') #blank plot, much bigger than plotting portion needed
# Set up axes
axis(side = 1, at = seqx, labels = seqx, cex.axis = 1, mgp = c(2, 1.5, 1)) # add axis and label (bottom)
mtext(side = 1, at = 0, 'Seizure Reduction', line = 2.5, cex = 0.85, padj = 1)
axis(side = 3, at = seqx, labels = seqx, cex.axis = 1, mgp = c(2, 1.5, 1)) # add axis and label (top)
mtext(side = 3, at = 0, 'Seizure Reduction', line = 2.5, cex = 0.85, padj = -1)
## add lines and dots
segments(df[, "ci_lo"], yy, df[,"ci_hi"], yy) # add lines
points(df[,"SR"]*100, yy, pch = 19) # add points
segments(x0 = 0, y0 = max(yy), y1 = 0, lty = 3, lwd = 0.75) #vertical line # 0
### Add text information
par(xpd = TRUE)
text(x = -550, y = yy, df[,"Study"], pos = 4)
text(x = -450, y = yy, df[,"SR"]*100, pos = 4)
text(x = -350, y = yy, df[,"SE"], pos = 4)
text(x = -250, y = yy, df[,"Patients"], pos = 4)
text(x = 150, y = yy, paste0(df[,"ci_lo"], "-", df[,"ci_hi"]), pos = 4)
text(x = c(seq(-550, -250, 100), 150), y = max(yy)+0.75,
c(colnames(df)[1:4], "CI"), pos = 4, font = 2)
# Add legend
legend(x = 50, y = 0.5, c("Point estimate", "95% Confidence interval"),
pch = c(19, NA), lty = c(NA, 19), bty = "n", cex = 0.65)
I am working with the R programming language. I am trying to plot some categorical and continuous data that I am working with, but I am getting an error that tells me that such plots are only possible with "only numeric variables".
library(survival)
library(ggplot2)
data(lung)
data = lung
data$sex = as.factor(data$sex)
data$status = as.factor(data$status)
data$ph.ecog = as.factor(data$ph.ecog)
str(data)
#plot
mycolours <- rainbow(length(unique(data$sex)), end = 0.6)
# png("gally.png", 500, 400, type = "cairo", pointsize = 14)
par(mar = c(4, 4, 0.5, 0.75))
plot(NULL, NULL, xlim = c(1, 5), ylim = range(data[, 1:6]) + c(-0.2, 0.2),
bty = "n", xaxt = "n", xlab = "Variable", ylab = "Standardised value")
axis(1, 1:5, labels = colnames(data)[1:6])
abline(v = 1:5, col = "#00000033", lwd = 2)
abline(h = seq(-2.5, 2.5, 0.5), col = "#00000022", lty = 2)
for (i in 1:nrow(data)) lines(as.numeric(data[i, 1:6]), col = mycolours[as.numeric(data$sex[i])])
legend("topright", c("Female", "Male"), lwd = 2, col = mycolours, bty = "n")
# dev.off()
Does anyone know if this is possible to do with both categorical and continuous data?
Thanks
Sources: R: Parallel Coordinates Plot without GGally
Yup. You just have to be careful with the values. Remember how the factors are coded internally: they are just spicy integer variables with value labels (similar to names). You can losslessly cast it to character or to numeric. For the sake of plotting, you need numbers for line coordinates, so the factor-y nature of your variables will come at the end.
Remember that the quality of your visualisation and the information content depends on the order of your variables in you data set. For factors, labels are absolutely necessary. Help the reader by doing some completely custom improvements impossible in ggplot2 in small steps!
I wrote a custom function allowing anyone to add super-legible text on top of the values that are not so obvious to interpret. Give meaningful names, choose appropriate font size, pass all those extra parameters to the custom function as an ellipsis (...)!
Here you can see that most of the dead patients are female and most of the censored ones are males. Maybe adding some points with slight jitter will give the reader idea about the distributions of these variables.
library(survival)
data(lung)
# Data preparation
lung.scaled <- apply(lung, 2, scale)
drop.column.index <- which(colnames(lung) == "sex")
lung.scaled <- lung.scaled[, -drop.column.index] # Dropping the split variable
split.var <- lung[, drop.column.index]
lung <- lung[, -drop.column.index]
mycolours <- rainbow(length(unique(split.var)), end = 0.6, v = 0.9, alpha = 0.4)
# png("gally.png", 500, 400, type = "cairo", pointsize = 14)
par(mar = c(5.5, 4, 0.5, 0.75))
plot(NULL, NULL, xlim = c(1, ncol(lung.scaled)), ylim = range(lung.scaled, na.rm = TRUE) + c(-0.2, 0.2),
bty = "n", xaxt = "n", xlab = "", ylab = "Standardised value")
axis(1, 1:ncol(lung.scaled), labels = colnames(lung), cex.axis = 0.95, las = 2)
abline(v = 1:ncol(lung), col = "#00000033", lwd = 2)
abline(h = seq(round(min(lung.scaled, na.rm = TRUE)), round(max(lung.scaled, na.rm = TRUE), 0.5)), col = "#00000022", lty = 2)
for (i in 1:nrow(lung.scaled)) lines(as.numeric(lung.scaled[i, ]), col = mycolours[as.numeric(split.var[i])])
legend("topleft", c("Female", "Male"), lwd = 3, col = mycolours, bty = "n")
# Labels for some categorical variables with a white halo for readability
labels.with.halo <- function(varname, data.scaled, labels, nhalo = 32, col.halo = "#FFFFFF44", hscale = 0.04, vscale = 0.04, ...) {
offsets <- cbind(cos(seq(0, 2*pi, length.out = nhalo + 1)) * hscale, sin(seq(0, 2*pi, length.out = nhalo + 1)) * vscale)[-(nhalo + 1), ]
ind <- which(colnames(data.scaled) == varname)
yvals <- sort(unique(data.scaled[, ind]))
for (i in 1:nhalo) text(rep(ind, length(yvals)) + offsets[i, 1], yvals + offsets[i, 2], labels = labels, col = col.halo, ...)
text(rep(ind, length(yvals)), yvals, labels = labels, ...)
}
labels.with.halo("status", lung.scaled, c("Censored", "Dead"), pos = 3)
labels.with.halo("ph.ecog", lung.scaled, c("Asymptomatic", "Symp. but ambul.", "< 50% bed", "> 50% bed"), pos = 3, cex = 0.9)
# dev.off()
I have the following code (Thanks to an answer from #Rawr in this question):
labes1 <- c("P(LNG)","","Volume(LNG)","","P(oil)","","Can.GDP","","US GDP","")
titles <- c("Levels","","","","","Log Difference","","","","")
par(mfrow = c(5, 2), mar = c(0.3, 6, 0, 2), oma = c(5, 0, 3, 2))
lapply(1:10, function(ii) {
x <- plotdata1[, ii, drop = FALSE]
plot(x, xlab = "Quarter", ylab = labes1[ii], axes = FALSE)
axis(2, las = 1)
box()
if (ii %in% 9:10) {
axis(1)
title(xlab = 'Quarter', xpd = NA)
}
if (ii %in% 1:2)
title(main = c('Levels', 'Log Difference')[ii], xpd = NA, line = 1)
})
This produces the following plot:
The obvious issue is the overlaying of the y-axis labels with the y-axis values. I have tried playing around with the mar() and oma() but these just change the margins around, I was hoping this would move things out of the way. How can I move the y-axis labels as separate from the plot? I will also be moving the margins a bit so that the white space between the two columns of plots will be closer together.
You can define the ylab separately, like what you're doing for the xlab, and set the line parameter to define its distance from the plot (as stated in this post).
I got a running example from combining your code and #rawr's from your previous question.
set.seed(1)
z <- ts(matrix(rt(200 * 10, df = 3), 200, 10), start = c(1961, 1), frequency = 12)
z <- z * 1e5 # to make "wide" y-axis labels
## vectors of x, y, and main labels
xl <- sprintf('x label %s', 1:10)
yl <- sprintf('y label %s', 1:10)
ml <- sprintf('main label %s', 1:10)
labes1 <- c("P(LNG)","","Volume(LNG)","","P(oil)","","Can.GDP","","US GDP","")
titles <- c("Levels","","","","","Log Difference","","","","")
par(mfrow = c(5, 2), mar = c(0.3, 6, 0, 2), oma = c(5, 0, 3, 2))
lapply(1:10, function(ii) {
x <- z[, ii, drop = FALSE]
plot(x, xlab = "Quarter", ylab = "", axes = FALSE) # set ylab to ""
axis(2, las = 1)
title(ylab = labes1[ii], line = 4) # set the line at an appropriate distance
box()
if (ii %in% 9:10) {
axis(1)
title(xlab = 'Quarter', xpd = NA)
}
if (ii %in% 1:2)
title(main = c('Levels', 'Log Difference')[ii], xpd = NA, line = 1)
})
The code above outputs the following graph for line = 4 :
and this plot for line = 3 :
I am having a 3D plot in which the points are colored acording to some extra vector. My problem is to add a color gradient legend. This is my code:
x = matrix(NA,100,6)
#x value
x[,1] = runif(100, 0, 10)
#y value
x[,2] = runif(100, 0, 10)
#z value
x[,3] = x[,1]+x[,2]
#additional value
x[,4] = runif(100, 0, 1)
#find out in which interval each additional value is
intervals = seq(0,1,1/10)
x[,5] = findInterval(x[,4], intervals)
colours = topo.colors(length(intervals))
x[,6] = colours[x[,5]]
library(rgl)
plot3d(as.numeric(x[,1]),as.numeric(x.stab.in[,2]), as.numeric(x[,3]),
type="p", col=x[,6], size=2, xlab = "x(t)", ylab = "y(t)",
zlab = "z(t)")
decorate3d(xlab = "x", ylab = "y", zlab = "z")
legend3d("topright", legend = intervals, pch = 16, col = colours, cex=1, inset=c(0.02))
grid3d(c("x", "y+", "z"),col = "gray")
The plot looks like this
but I want the legend in a gradient form. That means I don't want separate points for each color but one box in which the colors fade into each other.
Here is a possible solution if you are okay with using scatterplot3d package instead of rgl. It is basically same but non-interactive. Here is your code modified to produce your expected result.
x = matrix(NA,100,6)
#x value
x[,1] = runif(100, 0, 10)
#y value
x[,2] = runif(100, 0, 10)
#z value
x[,3] = x[,1]+x[,2]
#additional value
x[,4] = runif(100, 0, 1)
#find out in which interval each additional value is
intervals = seq(0,1,1/10)
x[,5] = findInterval(x[,4], intervals)
#produce gradient of colors
#you can define different colors (two or more)
gradient <- colorRampPalette(colors = c("yellow", "green", "blue"))
colours <- gradient(length(intervals))
x[,6] = colours[x[,5]]
library(scatterplot3d)
png('3d.png', width = 600, height = 400)
layout(matrix(1:2, ncol=2), width = c(3, 1), height = c(1, 1))
scatterplot3d(as.numeric(x[,1]),as.numeric(x[,2]), as.numeric(x[,3]), type = 'p',
cex.symbols = 1.25, color=x[,6], pch = 16, xlab = "x(t)", ylab = "y(t)", zlab = "z(t)")
plot(x = rep(1, 100), y = seq_along(x[,6]),
pch = 15, cex = 2.5,
col = gradient(length(x[,6])),
ann = F, axes = F, xlim = c(1, 2))
axis(side = 2, at = seq(1, nrow(x), length.out = 11),
labels = 1:11,
line = 0.15)
dev.off()
This will plot the following graph
Here is another solution if you want to plot a gradient on an interactive 3d plot, such as if you needed to animate the plot into a movie.
require(car)
require(rgl)
require(RColorBrewer)
require(mgcv)
require(magick) #Only for creating the animation of the plot as a gif
#Creating mock dataset
Example_Data <- data.frame(Axis1 = rnorm(100),
Axis2 = rnorm(100),
Axis3 = rnorm(100))
Example_Data$Value <- Example_Data$Axis1+Example_Data$Axis2
#Defining function that takes a vector of numeric values and converts them to
#a spectrum of rgb colors to help color my scatter3d plot
get_colors <- function(values){
v <- (values - min(values))/diff(range(values))
x <- colorRamp(rev(brewer.pal(11, "Spectral")))(v)
rgb(x[,1], x[,2], x[,3], maxColorValue = 255)
}
#Writing function that takes a vector of numeric values and a title and creates
#a gradient legend based on those values and the title and suitable for addition
#to a scatter3d plot via a call to bgplot3d()
#Note, I didn't have time to make this automatically adjust text position/size for different size
#plot windows, so values may need to be adjusted manually depending on the size of the plot window.
gradient_legend_3d <- function(values, title){
min_val <- min(values)
max_val <- max(values)
x <- colorRamp(brewer.pal(11, "Spectral"))((0:20)/20)
colors <- rgb(x[,1], x[,2], x[,3], maxColorValue = 255)
legend_image <- as.raster(matrix(colors, ncol=1))
plot(c(0,1),c(0,1),type = 'n', axes = F,xlab = '', ylab = '', main = '') #Generates a blank plot
text(x=0.92, y = seq(0.5, 1,l=5), labels = signif(seq(min_val, max_val,l=5), 2), cex = 1.5) #Creates the numeric labels on the scale
text(x = 0.85, y = 1, labels = title, adj = 1, srt = 90, cex = 1.5) #Determines where the title is placed
rasterImage(legend_image, 0.87, 0.5, 0.9,1) #Values can be modified here to alter where and how wide/tall the gradient is drawn in the plotting area
}
#Creating scatter3d plot
scatter3d(x = Example_Data$Axis1, y = Example_Data$Axis2, z = Example_Data$Axis3, xlab = "Axis1", ylab = "Axis2", zlab = "Axis3", surface = F, grid = F, ellipsoid = F, fogtype = "none", point.col = get_colors(Example_Data$Value))
#Changing size of plotting window and orientation to optimize for addition of static legend
#This may not work on another machine, so the window may need to be adjusted manually
par3d(windowRect = c(0,23,1536,824))
par3d(userMatrix = matrix(c(-0.98181450, -0.02413967, 0.18830180, 0, -0.03652956, 0.99736959, -0.06260729, 0, -0.18629514, -0.06834736, -0.98011345, 0, 0, 0, 0, 1), nrow = 4, ncol = 4, byrow = T))
#Adding legend
bgplot3d(gradient_legend_3d(Example_Data$Value, "Point Value"))
#Animating plot and saving as gif
movie3d(spin3d(axis = c(0,1,0), rpm = 5), duration = 12, dir = getwd(), fps = 5, convert = FALSE, clean = FALSE)
frames <- NULL
for(j in 0:60){
if(j == 1){
frames <- image_read(sprintf("%s%03d.png", "movie", j))
} else {
frames <- c(frames, image_read(sprintf("%s%03d.png", "movie", j)))
}
}
animation <- image_animate(frames, fps = 10, optimize = TRUE)
image_write(animation, path = "Example.gif")
for(j in 0:60){
unlink(sprintf("%s%03d.png", "movie", j))
}
See link to view 3d plot generated by this code:
gif of 3d plot with gradient color scale
I'm making plots like the one generated with the following code:
var1 <- sort(runif(10, 0, 1), decreasing = TRUE)
var2 <- sort(runif(10, 0, 1))
plot(var1, pch = 20, ylab = c("Var 1", "Var 2"))
points(var2, pch = 20, col = "grey")
Is there a way, with just the R graphics package, to place a black circle before Var 1 and a grey circle before Var 2 in the y axis label, to avoid having to insert a legend? Or alternatively, a way to use different text colours (black for Var 1 and grey for Var 2) in the y axis? I tried using col.lab = c("black","grey"), but it says Error in plot.window(...) : graphical parameter "col.lab" has the wrong length.
Many thanks in advance,
Márcia
I'm not sure how to add the point to the label, but an easy way to labe with color can be done in the following way:
var1 <- sort(runif(10, 0, 1), decreasing = TRUE)
var2 <- sort(runif(10, 0, 1))
plot(var1, pch = 20, ylab = "")
points(var2, pch = 20, col = "grey")
mtext("Var 1", side=2, line=2)
mtext("Var 2", side=2, line=3, col="grey")
Would something like this work for you? It's a bit busy on the left axis, but I think it shows what you are asking about.
> var1 <- sort(runif(10, 0, 1), decreasing = TRUE)
> var2 <- sort(runif(10, 0, 1))
> plot(var1, ylim = range(c(var1, var2)), pch = 20, ylab = "", axes = FALSE)
> points(var2, pch = 20, col = "grey")
> labs <- round(sort(c(var1, var2)), 1)
> axis(1)
> axis(2, at = sort(c(var1, var2)), labels = labs)
> sapply(var1, function(x) points(-0.1, x, pch = 20))
> sapply(var2, function(x) points(-0.1, x, pch = 20, col = "grey"))
> box()