This page shows how to create Sample.csv and Sample.tsv files using writeLines for use in later examples of read.csv and read.delim.
Find all the R code in this Gist, R-IO-writeLines-readLines.R.
Let's first define some data in a vector of five character strings:
data.lines.comma <- c("Grade,Age,Code,Amount,Start", "A,23,ABC,48.982,2014-01-01",
"B,56,DEF,2.7183,2014-01-02", "F,,,0,", "B,35,XYZ,3.1416,2014-01-05")
data.lines.comma
## [1] "Grade,Age,Code,Amount,Start" "A,23,ABC,48.982,2014-01-01"
## [3] "B,56,DEF,2.7183,2014-01-02" "F,,,0,"
## [5] "B,35,XYZ,3.1416,2014-01-05"
writeLines writes the strings to a comma-separated-value (CSV) text file:
writeLines(data.lines.comma, "Sample.csv")
Many ASCII editors can display a CSV file, such as the graphic below from using the gedit editor:
The R “global substitution” function, gsub, converts commas to tabs:
data.lines.tab <- gsub(",", "\t", data.lines.comma)
data.lines.tab
## [1] "Grade\tAge\tCode\tAmount\tStart" "A\t23\tABC\t48.982\t2014-01-01"
## [3] "B\t56\tDEF\t2.7183\t2014-01-02" "F\t\t\t0\t"
## [5] "B\t35\tXYZ\t3.1416\t2014-01-05"
writeLines can write these convered lines to a tab-separated-value (TSV) file:
writeLines(data.lines.tab, "Sample.tsv")
Note: Tab-delimited data are often written to .txt or .tab files, too.
Caution: Tabs provide ambiguous white space in different editors and editing them can be tricky.
The gedit editor only hints the file contains tabs by the left justification of columns, but the file could have spaces or tabs. Visual inspection does not identify whether a file has tabs or spaces or a mixture:
An ancient kedit editor can be configured to show tabs (and other special characters) as filled circles making it safe to edit the tabs:
The data can be read using readLines into a new R object and compared against the original vector of character strings:
data.lines.copy.comma <- readLines("Sample.csv")
data.lines.copy.comma
## [1] "Grade,Age,Code,Amount,Start" "A,23,ABC,48.982,2014-01-01"
## [3] "B,56,DEF,2.7183,2014-01-02" "F,,,0,"
## [5] "B,35,XYZ,3.1416,2014-01-05"
# Compare with original
data.lines.copy.comma == data.lines.comma
## [1] TRUE TRUE TRUE TRUE TRUE
all(data.lines.copy.comma == data.lines.comma)
## [1] TRUE
Read tsv from disk to verify write was successful
data.lines.copy.tab <- readLines("Sample.tsv")
data.lines.copy.tab
## [1] "Grade\tAge\tCode\tAmount\tStart" "A\t23\tABC\t48.982\t2014-01-01"
## [3] "B\t56\tDEF\t2.7183\t2014-01-02" "F\t\t\t0\t"
## [5] "B\t35\tXYZ\t3.1416\t2014-01-05"
# Compare with original
data.lines.copy.tab == data.lines.tab
## [1] TRUE TRUE TRUE TRUE TRUE
all(data.lines.copy.tab == data.lines.tab)
## [1] TRUE
Let's save a copy of The R Journal archive page for future comparison to detect what changes have been made.
s <- readLines("http://journal.r-project.org/archive/")
length(s)
## [1] 256
head(s)
## [1] "<!DOCTYPE html>"
## [2] ""
## [3] "<html>"
## [4] "<head>"
## [5] " <title>Past issues. The R Journal</title>"
## [6] " <link rel=\"stylesheet\" type=\"text/css\" href=\"/r-journal.css\">"
tail(s)
## [1] ""
## [2] " ga('create', 'UA-40966673-1', 'r-project.org');"
## [3] " ga('send', 'pageview');"
## [4] "</script>"
## [5] "</body>"
## [6] "</html>"
tail(head(s, 65), 1) # Current issue (for now)
## [1] " <li><a href=\"2013-2/\">2013. Issue 2, Dec</a></li> "
# Save copy locally to look for differences in future
timestamp <- format(Sys.time(), "%Y-%m-%d-%H%M%S")
filename <- paste0("R-Journal-Archive-page-", timestamp, ".html")
filename
## [1] "R-Journal-Archive-page-2014-01-15-232337.html"
writeLines(s, filename)