我想修改phyloseq包生成的图(从github下载)。 Phyloseq图是ggplot2对象,所以我认为我可以通过将ggplot2对象添加到phyloseq创建的对象来添加元素。在某些情况下,这可行,但在其他情况下则不然,我不明白为什么。例如:
require(phyloseq)
require(grid)
require(ggplot2)
require(plyr)
#use the GlobalPatterns dataset from the Phyloseq package
GP <- GlobalPatterns
#do some preprocessing to the data
wh0 <- genefilter_sample(GP, filterfun_sample(function(x) x > 5), A = 0.5 * nsamples(GP))
GP1 <- prune_taxa(wh0, GP)
GP1 <- transform_sample_counts(GP1, function(x) 1e+06 * x/sum(x))
phylum.sum = tapply(taxa_sums(GP1), tax_table(GP1)[, "Phylum"], sum, na.rm = TRUE)
top5phyla = names(sort(phylum.sum, TRUE))[1:5]
GP1 <- prune_taxa((tax_table(GP1)[, "Phylum"] %in% top5phyla), GP1)
#ordination for NMDS plot using a Bray-Curtis distance
GP.ord <- ordinate(GP1, "NMDS", "bray")
#create plot
p3 <- plot_ordination(GP1, GP.ord, type = "biplot", color = "SampleType", shape = "Phylum", title = "biplot")
现在我将尝试从素食包中添加一些envfit()箭头,参见上一个问题here:
require(vegan)
# First, lets apply envfit to the human/not human variable
human = get_variable(GP1, "SampleType") %in% c("Feces", "Mock", "Skin", "Tongue")
sample_data(GP1)$human <- factor(human)
nmds.envfit <- envfit(GP.ord$points, env = as.data.frame(sample_data(GP1)$human), perm = 999) #standard envfit
str(nmds.envfit)
#data for the envfit arrows
vec.sp.df<-as.data.frame(cbind((nmds.envfit$factors$centroids*sqrt(nmds.envfit$factors$r)),pvals=nmds.envfit$factors$pvals)) #this is necessary, see Gavin Simpson in the link provided above
env.scores.nmds <- as.data.frame(vec.sp.df[vec.sp.df$pvals<0.05,]) #extracts relevant scores from envifit
#extracts relevant scores from envifit
env.scores.nmds <- cbind(env.scores.nmds, env.variables = c("Not Human", "Human")) #and then gives them their names
env.scores.nmds
mult<- 1 #can change this if the arrows need to be a different length
###Now let us add these vectors to p3
p3 + geom_segment(data = env.scores.nmds,
aes(x = 0, xend = mult*MDS1, y = 0, yend = mult*MDS2),
arrow = arrow(length = unit(0.75, "cm")), colour = "black") + #arrows for envfit. doubled the length for similarity to the plot() function. NB check ?envfit regarding arrow length if not familiar with lengths
geom_text(data = env.scores.nmds, #labels the environmental variable arrows * "mult" as for the arrows
aes(x = mult*MDS1, y = mult*MDS2, label=env.variables),
size = 6,
hjust = -0.5)
但是,这会返回错误: “eval中的错误(expr,envir,enclos):找不到对象'id.type'”
如果我们尝试添加另一种类型的ggplot2元素,它将起作用:
p3+ geom_hline(yintercept=0.75)
答案 0 :(得分:0)
错误消息已经让您知道在添加的图层中需要修复的内容。
要修改的ggplot2对象在p3$data
或其中一个图层的$data
插槽中有一个列变量id.type
,这是一个美学映射参数如果您没有覆盖它,则会隐式传递给您的新图层。鉴于您添加的图层在两种情况下都指定了x和y,我怀疑inherit.aes=FALSE
是一个分面或颜色变量。在ggplot2的最新版本中,您可以包含参数id.type
,以避免这种继承的映射,在这种情况下,您将缺少未指定的映射。结果是不同的取决于它是什么(例如,如果facet,然后层,在我看来两个面板中出现;如果是颜色,那么图层将被指定为默认颜色。)
或者,您可以为新图层的数据添加{{1}}列。这取决于你想要达到的结果。