用R中的grep删除文本文件中的行
我有一个非常重复的文本文件,前几行是这样的:用R中的grep删除文本文件中的行,r,dataframe,data-cleaning,R,Dataframe,Data Cleaning,我有一个非常重复的文本文件,前几行是这样的: Filename: ROI: red_1 [Red] 20 points Basic Stats Min Max Mean Stdev Band 1 0.013282 0.133982 0.061581 0.034069 Band 2 0.009866 0.112935 0.042688 0.026618 Band 3 0.008
Filename:
ROI: red_1 [Red] 20 points
Basic Stats Min Max Mean Stdev
Band 1 0.013282 0.133982 0.061581 0.034069
Band 2 0.009866 0.112935 0.042688 0.026618
Band 3 0.008304 0.037059 0.018434 0.007515
Band 4 0.004726 0.040089 0.018490 0.009605
Histogram DN Npts Total Percent Acc Pct
Band 1 0.013282 1 1 5.0000 5.0000
Bin=0.00047 0.013755 0 1 0.0000 5.0000
0.014228 0 1 0.0000 5.0000
。。并持续一段时间,直到达到另一个ROI值,如下所示:
Stats for ROI: red_5 [Red] 20 points
Basic Stats Min Max Mean Stdev
Band 1 0.030513 0.180980 0.090056 0.044456
Band 2 0.022289 0.157861 0.046419 0.030555
Band 3 0.012533 0.046693 0.027343 0.008947
Band 4 0.003332 0.041555 0.016888 0.007770
Histogram DN Npts Total Percent Acc Pct
Band 1 0.030513 1 1 5.0000 5.0000
Bin=0.00059 0.031103 0 1 0.0000 5.0000
0.031693 0 1 0.0000 5.0000
0.032283 0 1 0.0000 5.0000
ROI Band min max mean stdev
red_1 [Red] 20 points Band 1 0.013282 0.133982 0.061581 0.034069
red_1 [Red] 20 points Band 2 0.009866 0.112935 0.042688 0.026618
red_1 [Red] 20 points Band 3 0.008304 0.037059 0.018434 0.007515
red_1 [Red] 20 points Band 4 0.004726 0.040089 0.018490 0.009605
red_2 [Red] 12 points Band 1 0.032262 0.124425 0.078073 0.028031
red_2 [Red] 12 points Band 2 0.021072 0.064156 0.037923 0.012178
red_2 [Red] 12 points Band 3 0.013404 0.066043 0.036316 0.014787
red_2 [Red] 12 points Band 4 0.005162 0.055781 0.015526 0.013255
red_3 [Red] 12 points Band 1 0.037488 0.107830 0.057892 0.018964
red_3 [Red] 12 points Band 2 0.028140 0.072370 0.045340 0.014507
我只想要一个数据框,其中只有ROI:…,基本统计数据,第1、2栏。。每个投资回报率为4级。对于所有100个ROI,最终输出将类似于此
ROI: red_1 [Red] 20 points
Basic Stats Min Max Mean Stdev
Band 1 0.013282 0.133982 0.061581 0.034069
Band 2 0.009866 0.112935 0.042688 0.026618
Band 3 0.008304 0.037059 0.018434 0.007515
Band 4 0.004726 0.040089 0.018490 0.009605
Stats for ROI: red_5 [Red] 20 points
Basic Stats Min Max Mean Stdev
Band 1 0.030513 0.180980 0.090056 0.044456
Band 2 0.022289 0.157861 0.046419 0.030555
Band 3 0.012533 0.046693 0.027343 0.008947
Band 4 0.003332 0.041555 0.016888 0.007770
有人能给我指一个在R中清理数据的教程吗?使用repgrepl是正确的方法吗?此处提供了数据的全文文件:。如注释中所述,为每个ROI值设置换行符不适合R数据框的形式。我认为我们可以通过添加一个列来跟踪ROI来实现您所期望的目标。最终产品如下所示:
Stats for ROI: red_5 [Red] 20 points
Basic Stats Min Max Mean Stdev
Band 1 0.030513 0.180980 0.090056 0.044456
Band 2 0.022289 0.157861 0.046419 0.030555
Band 3 0.012533 0.046693 0.027343 0.008947
Band 4 0.003332 0.041555 0.016888 0.007770
Histogram DN Npts Total Percent Acc Pct
Band 1 0.030513 1 1 5.0000 5.0000
Bin=0.00059 0.031103 0 1 0.0000 5.0000
0.031693 0 1 0.0000 5.0000
0.032283 0 1 0.0000 5.0000
ROI Band min max mean stdev
red_1 [Red] 20 points Band 1 0.013282 0.133982 0.061581 0.034069
red_1 [Red] 20 points Band 2 0.009866 0.112935 0.042688 0.026618
red_1 [Red] 20 points Band 3 0.008304 0.037059 0.018434 0.007515
red_1 [Red] 20 points Band 4 0.004726 0.040089 0.018490 0.009605
red_2 [Red] 12 points Band 1 0.032262 0.124425 0.078073 0.028031
red_2 [Red] 12 points Band 2 0.021072 0.064156 0.037923 0.012178
red_2 [Red] 12 points Band 3 0.013404 0.066043 0.036316 0.014787
red_2 [Red] 12 points Band 4 0.005162 0.055781 0.015526 0.013255
red_3 [Red] 12 points Band 1 0.037488 0.107830 0.057892 0.018964
red_3 [Red] 12 points Band 2 0.028140 0.072370 0.045340 0.014507
由于您的数据并不庞大,因此我们可以通过将数据作为字符向量读入并遍历以下行来实现这一点:
file <- file("test2.txt",)
dat <- readLines(file)
out <- NULL
roi <- NULL
for(i in 1:length(dat)){
line <- dat[i]
if(length(grep("ROI: ",line))>0){
roi <- substr(line,regexpr("ROI",line)[1]+5,nchar(line))
}
if(substr(line,0,9)==" Band"){
splitLine <- strsplit(trimws(line),"\t")[[1]]
outLine <- data.frame("ROI" = roi,
"Band" = splitLine[1],
"min" = splitLine[2],
"max" = splitLine[3],
"mean" = splitLine[4],
"stdev" = splitLine[5]
)
out <- rbind(out,outLine)
}
}
数据框应该非常接近您要查找的内容。如评论中所述,为每个ROI值设置换行符不适合R数据框的形式。我认为我们可以通过添加一个列来跟踪ROI来实现您所期望的目标。最终产品如下所示:
Stats for ROI: red_5 [Red] 20 points
Basic Stats Min Max Mean Stdev
Band 1 0.030513 0.180980 0.090056 0.044456
Band 2 0.022289 0.157861 0.046419 0.030555
Band 3 0.012533 0.046693 0.027343 0.008947
Band 4 0.003332 0.041555 0.016888 0.007770
Histogram DN Npts Total Percent Acc Pct
Band 1 0.030513 1 1 5.0000 5.0000
Bin=0.00059 0.031103 0 1 0.0000 5.0000
0.031693 0 1 0.0000 5.0000
0.032283 0 1 0.0000 5.0000
ROI Band min max mean stdev
red_1 [Red] 20 points Band 1 0.013282 0.133982 0.061581 0.034069
red_1 [Red] 20 points Band 2 0.009866 0.112935 0.042688 0.026618
red_1 [Red] 20 points Band 3 0.008304 0.037059 0.018434 0.007515
red_1 [Red] 20 points Band 4 0.004726 0.040089 0.018490 0.009605
red_2 [Red] 12 points Band 1 0.032262 0.124425 0.078073 0.028031
red_2 [Red] 12 points Band 2 0.021072 0.064156 0.037923 0.012178
red_2 [Red] 12 points Band 3 0.013404 0.066043 0.036316 0.014787
red_2 [Red] 12 points Band 4 0.005162 0.055781 0.015526 0.013255
red_3 [Red] 12 points Band 1 0.037488 0.107830 0.057892 0.018964
red_3 [Red] 12 points Band 2 0.028140 0.072370 0.045340 0.014507
由于您的数据并不庞大,因此我们可以通过将数据作为字符向量读入并遍历以下行来实现这一点:
file <- file("test2.txt",)
dat <- readLines(file)
out <- NULL
roi <- NULL
for(i in 1:length(dat)){
line <- dat[i]
if(length(grep("ROI: ",line))>0){
roi <- substr(line,regexpr("ROI",line)[1]+5,nchar(line))
}
if(substr(line,0,9)==" Band"){
splitLine <- strsplit(trimws(line),"\t")[[1]]
outLine <- data.frame("ROI" = roi,
"Band" = splitLine[1],
"min" = splitLine[2],
"max" = splitLine[3],
"mean" = splitLine[4],
"stdev" = splitLine[5]
)
out <- rbind(out,outLine)
}
}
dataframe out应该非常接近您要查找的内容。这里有一个方法,可以将文件作为文本拉入,识别ROI的相关行和基本统计行,然后生成一个包含所有数据的长数据框
input <- readLines("https://dl.dropboxusercontent.com/u/45095175/test2.txt")
roi_lines <- grep("ROI", input)
basic_stat_lines <- grep("Basic Stats", input)
roi_names <- sub("^.*ROI: ", "", input[roi_lines])
roi_data <- lapply(1:length(basic_stat_lines), function(i) {
data.frame(roi = roi_names[i], read.delim(text = input[basic_stat_lines[i] + 0:4],
stringsAsFactors = FALSE, check.names = FALSE),
stringsAsFactors = FALSE)
})
roi_data_all <- do.call("rbind", roi_data)
这里有一种方法,它将文件作为文本拉入,识别ROI的相关行和基本统计行,然后生成包含所有数据的长数据帧
input <- readLines("https://dl.dropboxusercontent.com/u/45095175/test2.txt")
roi_lines <- grep("ROI", input)
basic_stat_lines <- grep("Basic Stats", input)
roi_names <- sub("^.*ROI: ", "", input[roi_lines])
roi_data <- lapply(1:length(basic_stat_lines), function(i) {
data.frame(roi = roi_names[i], read.delim(text = input[basic_stat_lines[i] + 0:4],
stringsAsFactors = FALSE, check.names = FALSE),
stringsAsFactors = FALSE)
})
roi_data_all <- do.call("rbind", roi_data)
myfun将文本文件作为输入,并返回数据帧列表。什么参数将设置用户是否要从文本文件中提取数据或基本统计信息
myfun <- function( file, what )
{
x <- readLines( file )
g1 <- which( grepl("ROI:", x))
if( what == 'Basic Stats'){
g2 <- which( grepl('Basic Stats', x))
} else if ( what == "Histogram" ) {
g2 <- which( grepl("Histogram", x))
} else {
stop( 'what value is not supported')
}
df_list <- list()
counter <- 0
while( counter < length( g1 ))
{
counter <- counter + 1
if( counter != length( g1 ) ){
low <- g1[ counter ]
high <- g1[ counter + 1 ]
} else {
low <- g1[ counter ]
high <- length( x )
}
min_ind <- min( g2[ which( g2 > low & g2 < high ) ] )
title <- ifelse( counter == 1,
list( gsub( '\\[|\\]', '', unlist( strsplit( x[ low ], "\ ") )[ 2:4 ] ) ),
list( gsub( '\\[|\\]', '', unlist( strsplit( x[ low ], "\ ") )[ 4:6 ] ) ) )
if( what == 'Basic Stats'){
min_ind <- min( g2[ which( g2 > low & g2 < high ) ] )
x1 <- data.frame( do.call( 'rbind', strsplit( x[ min_ind: ( min_ind + 5 ) ], "\t")), stringsAsFactors = FALSE )
colnames( x1 ) <- x1[1, ]
x1 <- x1[2:5, ]
x1 <- do.call( 'cbind', list( x1, do.call( 'rbind', title )))
colnames(x1)[(ncol(x1)-2): ncol(x1)] <- c( 'ROI', 'color', 'points') # column names of last 3 columns
colnames(x1) <- gsub("\ ", '', colnames(x1)) # remove spaces
# convert from character to numeric data type
x1[, 2:5 ] <- lapply( x1[, 2:5 ], function(x) as.numeric( as.character( x ) ) )
df_list[[ as.character(counter) ]] <- x1
} else if ( what == "Histogram" ) {
x1 <- data.frame( do.call( 'rbind', strsplit( x[ min_ind: (high-1) ], "\t")), stringsAsFactors = FALSE )
# column names and band and bin columns
colnames( x1 ) <- x1[1, ]
colnames(x1)[1] <- 'Histogram'
x1$Band <- rep( gsub("[Band\ ]", '', grep( "Band", x1$Histogram, value = TRUE )),
diff( c( grep( "Histogram", x1$Histogram ), ( nrow(x1) + 1 ) ) ) )
x1$Bin <- rep( gsub("[Bin=\ ]", '', grep( "Bin", x1$Histogram, value = TRUE )),
diff( c( grep( "Histogram", x1$Histogram ), ( nrow(x1) + 1 ) ) ) )
x1 <- x1[! grepl( 'Histogram', x1$Histogram ), ]
x1$Histogram <- NULL
x1 <- do.call( 'cbind', list( x1, do.call( 'rbind', title )))
colnames(x1)[(ncol(x1)-2): ncol(x1)] <- c( 'ROI', 'color', 'points') # column names of last 3 columns
colnames(x1) <- gsub("\ ", '', colnames(x1)) # remove spaces
# convert from character to numeric data type
x1[, c(1:7, 10) ] <- lapply( x1[, c(1:7, 10) ], function(x) as.numeric( as.character( x ) ) )
df_list[[ as.character(counter) ]] <- x1
}
}
return( df_list )
}
统计摘要:
myfun将文本文件作为输入,并返回数据帧列表。什么参数将设置用户是否要从文本文件中提取数据或基本统计信息
myfun <- function( file, what )
{
x <- readLines( file )
g1 <- which( grepl("ROI:", x))
if( what == 'Basic Stats'){
g2 <- which( grepl('Basic Stats', x))
} else if ( what == "Histogram" ) {
g2 <- which( grepl("Histogram", x))
} else {
stop( 'what value is not supported')
}
df_list <- list()
counter <- 0
while( counter < length( g1 ))
{
counter <- counter + 1
if( counter != length( g1 ) ){
low <- g1[ counter ]
high <- g1[ counter + 1 ]
} else {
low <- g1[ counter ]
high <- length( x )
}
min_ind <- min( g2[ which( g2 > low & g2 < high ) ] )
title <- ifelse( counter == 1,
list( gsub( '\\[|\\]', '', unlist( strsplit( x[ low ], "\ ") )[ 2:4 ] ) ),
list( gsub( '\\[|\\]', '', unlist( strsplit( x[ low ], "\ ") )[ 4:6 ] ) ) )
if( what == 'Basic Stats'){
min_ind <- min( g2[ which( g2 > low & g2 < high ) ] )
x1 <- data.frame( do.call( 'rbind', strsplit( x[ min_ind: ( min_ind + 5 ) ], "\t")), stringsAsFactors = FALSE )
colnames( x1 ) <- x1[1, ]
x1 <- x1[2:5, ]
x1 <- do.call( 'cbind', list( x1, do.call( 'rbind', title )))
colnames(x1)[(ncol(x1)-2): ncol(x1)] <- c( 'ROI', 'color', 'points') # column names of last 3 columns
colnames(x1) <- gsub("\ ", '', colnames(x1)) # remove spaces
# convert from character to numeric data type
x1[, 2:5 ] <- lapply( x1[, 2:5 ], function(x) as.numeric( as.character( x ) ) )
df_list[[ as.character(counter) ]] <- x1
} else if ( what == "Histogram" ) {
x1 <- data.frame( do.call( 'rbind', strsplit( x[ min_ind: (high-1) ], "\t")), stringsAsFactors = FALSE )
# column names and band and bin columns
colnames( x1 ) <- x1[1, ]
colnames(x1)[1] <- 'Histogram'
x1$Band <- rep( gsub("[Band\ ]", '', grep( "Band", x1$Histogram, value = TRUE )),
diff( c( grep( "Histogram", x1$Histogram ), ( nrow(x1) + 1 ) ) ) )
x1$Bin <- rep( gsub("[Bin=\ ]", '', grep( "Bin", x1$Histogram, value = TRUE )),
diff( c( grep( "Histogram", x1$Histogram ), ( nrow(x1) + 1 ) ) ) )
x1 <- x1[! grepl( 'Histogram', x1$Histogram ), ]
x1$Histogram <- NULL
x1 <- do.call( 'cbind', list( x1, do.call( 'rbind', title )))
colnames(x1)[(ncol(x1)-2): ncol(x1)] <- c( 'ROI', 'color', 'points') # column names of last 3 columns
colnames(x1) <- gsub("\ ", '', colnames(x1)) # remove spaces
# convert from character to numeric data type
x1[, c(1:7, 10) ] <- lapply( x1[, c(1:7, 10) ], function(x) as.numeric( as.character( x ) ) )
df_list[[ as.character(counter) ]] <- x1
}
}
return( df_list )
}
统计摘要:
这个问题没有得到足够的重视,只是因为您没有费心添加语言标记。您的问题是关于处理文本的,但您稍后会要求使用该数据创建data.frame。在一个DATA框架中间有ROI统计数据对我来说是没有意义的。请手动制作一个data.frame,显示前两组基本统计数据和波段行的组合方式好吗?这个问题没有得到足够的重视,只是因为您没有添加语言标记。您的问题是关于处理文本的,但稍后您会要求使用该数据创建data.frame。在一个DATA框架中间有ROI统计数据对我来说是没有意义的。你能手动制作一个data.frame,显示前两组基本统计数据和波段行的组合方式吗?