Question

我正在尝试生成一个质量控制报告，该报告在几个文件夹上循环（sapply）（每个文件夹对应一个实验），并且对于每个加载结果，创建表和图（在函数内）。生成的pdf应该包含文件夹的名称，然后按顺序包含表格和图表。我首先创建了R脚本（运行良好），然后创建了一个rnw文件。确实生成了这些图，但有2个问题（pdf outpout）：

loop_n_plots没有生成表格;
创建完所有图后，会出现一条看似列表输出的无意混乱线。

问：如何获取pdf中的表格？在块“table_files”中生成的表有效，但apply函数内部的表没有。为什么？更一般地说，我正在尝试做什么（我是怎么做的）对knitr报告好吗？是否最好在列表中添加表格和图表，然后循环列表以打印它们？

我现在已经使用了大块设置，但没有任何效果。

示例代码：

\documentclass{report}

\begin{document}
\title{Sequencing Quality Report}
\author{Deep Sequencing Group - SFB655}
\maketitle


<<knitr_option, cache=FALSE, echo=FALSE, results='hide'>>=
library(knitr)
## set global chunk options
opts_chunk$set(fig.align='center', fig.width=14, fig.heigth=8, out.width="1.2\\textwidth",  par=TRUE)
@


<<R_arguments, cache=FALSE, echo=FALSE, include=FALSE>>=

###### Libraries ######
library(reshape)
library(ggplot2)
theme_set(theme_bw(16)) # removes grey grid and increases letter size. Ideal for presentations
library(RColorBrewer)
library(plyr)
library(scales) # for natural numbers in axis
library(xtable)
library(rattle) # needed to generate a table in knitr?
#######################


###### Function definitions ######
## ggplot theme with extra space between legends and axis
gg.axis.space <- theme(axis.title.y=element_text(vjust=0.2), axis.title.x=element_text(vjust=0.2))


ReturnStatsPlotsAndTables <- function(fqc.folder){

   # for(fqc.folder in fq_fastqc.folders){
   ######################################
   ## for each folder in the vector will
   ## plot stats and 
   ## print tables of fastQC results

   ## which library is being analysed?
   fastq.lib <- data.frame(Libraries = gsub(".*/(L.*)\\.fq_fastqc", "\\1", fqc.folder, perl=T))
   xtable(fastq.lib)

   ## Basic statistics - table ##
   stats.path <- paste(fqc.folder, "/", "Basic_Statistics_fastqc_data.temp", sep="")
   basic.stats <- read.table(stats.path, header = TRUE, sep = "\t", stringsAsFactors = FALSE)
#    basic.stats[ ,1:2]
   xtable(basic.stats[ ,1:2]) 


   ## Summary of filters - table ##
   stats.path <- paste(fqc.folder, "/", "filters_summary_fastqc_data.temp", sep="")
   summary.filters <-  read.table(stats.path, 
      header = TRUE, sep = "\t", stringsAsFactors = FALSE)
#    summary.filters
   xtable(summary.filters)


   ## Per base sequence quality ##
   stats.path <- paste(fqc.folder, "/", "Per_base_sequence_quality_fastqc_data.temp", sep="")
   base.qual <- read.table(stats.path, 
      header = TRUE, sep = "\t", stringsAsFactors = FALSE)


   base.qual$Base <- factor(base.qual$Base, as.character(base.qual$Base)) # re-order the levels by order of appearance in DF

   plot.new()
   base.qual.p <- ggplot(base.qual, aes(x = Base, ymin = X10th.Percentile, lower = Lower.Quartile, middle = Median, upper = Upper.Quartile, ymax = X90th.Percentile, fill = Lower.Quartile)) + geom_boxplot(stat = "identity") + 
      theme(axis.text.x = element_text(angle=30, hjust=1, vjust=1)) + 
      annotate("rect", xmin=-Inf, xmax=Inf, ymin=0, ymax=20, alpha=0.1, fill="red") +
      annotate("rect", xmin=-Inf, xmax=Inf, ymin=20, ymax=28, alpha=0.1, fill="yellow") +
      annotate("rect", xmin=-Inf, xmax=Inf, ymin=28, ymax=Inf, alpha=0.1, fill="green") +
      ggtitle("Per base sequence quality") + ylab("Quality score (Phred score) ") + xlab("Position of base in read")

   print(base.qual.p)

}
@

\chapter{Preamble}

This an automated quality control report generated for the following fastq files:

<<table_files, echo=FALSE, results="asis">>=
##############################################
## loop over fastQC folder and parse txt files:

## list and read fastqc_data.temp old files
# testing #
# setwd("/projects/seq-work/analysis/martinad/p0196-totalRNA/")
folder <- "./"
filenames <- list.files(path=folder, pattern="fastqc_data.temp", recursive=TRUE) 
fq_fastqc.folders <- unique(dirname(filenames)) # the folders that contain fastQC
fastq.libs <- data.frame(Libraries = gsub(".*/(L.*)\\.fq_fastqc", "\\1", fq_fastqc.folders, perl=T))
xtable(fastq.libs)
@



\chapter{FastQC}

<<loop_n_plots, echo=FALSE, results='asis'>>=
## do the plotting
sapply(fq_fastqc.folders[1:3], ReturnStatsPlotsAndTables)
@

\end{document}

函数ReturnStatsPlotsAndTables实际上更长，这足以让我们知道发生了什么。

Answer 1

找到有两个步骤的解决方案：

用包含函数ReturnStatsPlotsAndTables的指令的for循环替换sapply;
需要使用以下方式显式打印表：

打印（xtable（fastq.lib））

以下是最终代码：

\documentclass{report}

\begin{document}
\title{Sequencing Quality Report}
\author{Deep Sequencing Group - SFB655}
\maketitle


<<knitr_option, cache=FALSE, echo=FALSE, results='hide'>>=
library(knitr)
## set global chunk options
opts_chunk$set(fig.align='center', fig.width=14, fig.heigth=8, out.width="1.2\\textwidth",  par=TRUE)
@


<<R_arguments, cache=FALSE, echo=FALSE, include=FALSE>>=

###### Libraries ######
library(reshape)
library(ggplot2)
theme_set(theme_bw(16)) # removes grey grid and increases letter size. Ideal for presentations
library(RColorBrewer)
library(plyr)
library(scales) # for natural numbers in axis
library(xtable)
library(rattle) # needed to generate a table in knitr?
#######################


###### Function definitions ######
## ggplot theme with extra space between legends and axis
gg.axis.space <- theme(axis.title.y=element_text(vjust=0.2), axis.title.x=element_text(vjust=0.2))

@


\chapter{Preamble}

This an automated quality control report generated for the following fastq files:

<<table_files, echo=FALSE, results="asis">>=
##############################################
## loop over fastQC folder and parse txt files:

## list and read fastqc_data.temp old files
# testing #
# setwd("/projects/seq-work/analysis/martinad/p0196-totalRNA/")
folder <- "./"
filenames <- list.files(path=folder, pattern="fastqc_data.temp", recursive=TRUE) 
fq_fastqc.folders <- unique(dirname(filenames)) # the folders that contain fastQC
fastq.libs <- data.frame(Libraries = gsub(".*/(L.*)\\.fq_fastqc", "\\1", fq_fastqc.folders, perl=T))
xtable(fastq.libs)
@



\chapter{FastQC}

<<loop_n_plots, echo=FALSE, results="asis">>=
## do the plotting
# sapply(fq_fastqc.folders[1:3], ReturnStatsPlotsAndTables)
for (fqc.folder in fq_fastqc.folders[1:2]){
   # for(fqc.folder in fq_fastqc.folders){
   ######################################
   ## for each folder in the vector will
   ## plot stats and 
   ## print tables of fastQC results
#    print(fqc.folder)
   ## which library is being analysed?
   fastq.lib <- data.frame(Libraries = gsub(".*/(L.*)\\.fq_fastqc", "\\1", fqc.folder, perl=T))
   print(xtable(fastq.lib))

   ## Basic statistics - table ##
   stats.path <- paste(fqc.folder, "/", "Basic_Statistics_fastqc_data.temp", sep="")
   basic.stats <- read.table(stats.path, header = TRUE, sep = "\t", stringsAsFactors = FALSE)
#    basic.stats[ ,1:2]
   print(xtable(basic.stats[ ,1:2])) 


   ## Summary of filters - table ##
   stats.path <- paste(fqc.folder, "/", "filters_summary_fastqc_data.temp", sep="")
   summary.filters <-  read.table(stats.path, 
      header = TRUE, sep = "\t", stringsAsFactors = FALSE)
#    summary.filters
   print(xtable(summary.filters))


   ## Per base sequence quality ##
   stats.path <- paste(fqc.folder, "/", "Per_base_sequence_quality_fastqc_data.temp", sep="")
   base.qual <- read.table(stats.path, 
      header = TRUE, sep = "\t", stringsAsFactors = FALSE)


   base.qual$Base <- factor(base.qual$Base, as.character(base.qual$Base)) # re-order the levels by order of appearance in DF

   plot.new()
   base.qual.p <- ggplot(base.qual, aes(x = Base, ymin = X10th.Percentile, lower = Lower.Quartile, middle = Median, upper = Upper.Quartile, ymax = X90th.Percentile, fill = Lower.Quartile)) + geom_boxplot(stat = "identity") + 
      theme(axis.text.x = element_text(angle=30, hjust=1, vjust=1)) + 
      annotate("rect", xmin=-Inf, xmax=Inf, ymin=0, ymax=20, alpha=0.1, fill="red") +
      annotate("rect", xmin=-Inf, xmax=Inf, ymin=20, ymax=28, alpha=0.1, fill="yellow") +
      annotate("rect", xmin=-Inf, xmax=Inf, ymin=28, ymax=Inf, alpha=0.1, fill="green") +
      ggtitle("Per base sequence quality") + ylab("Quality score (Phred score) ") + xlab("Position of base in read")

   print(base.qual.p)

}
@

\end{document}

如何创建循环文件夹并返回表格和图表的knitr报告

1 个答案: