help(read.table)13 File input-output
Reading data from a text file
Data is frequently stored in tabular form in text files. The read.table() function can read a file from your disk and return a data frame containing that data.
Suppose we have a data file mydata.txt with the following contents:
Can 1.70 65
Cem 1.75 66
Hande 1.62 61
Lale 1.76 64
Arda 1.78 63
Bilgin 1.77 84
Cem 1.69 75
Ozlem 1.75 65
Ali 1.73 75
Haluk 1.71 81The file can be read into a data frame simply with:
hwdata <- read.table("mydata.txt")
hwdata V1 V2 V3
1 Can 1.70 65
2 Cem 1.75 66
3 Hande 1.62 61
4 Lale 1.76 64
5 Arda 1.78 63
6 Bilgin 1.77 84
7 Cem 1.69 75
8 Ozlem 1.75 65
9 Ali 1.73 75
10 Haluk 1.71 81class(hwdata)[1] "data.frame"We can change the columns of the dataframe as usual:
names(hwdata) <- c("Name", "Height","Weight")
hwdata Name Height Weight
1 Can 1.70 65
2 Cem 1.75 66
3 Hande 1.62 61
4 Lale 1.76 64
5 Arda 1.78 63
6 Bilgin 1.77 84
7 Cem 1.69 75
8 Ozlem 1.75 65
9 Ali 1.73 75
10 Haluk 1.71 81The function read.table() is quite versatile, and it has a lot of parameters to tune its behavior. The help documentation help(read.table) can be helpful.
Let’s read a new data file mydata2.txt. It has a header row, and we want to set the column names of the resulting data frame accordingly:
Name Height Weight
Can 1.70 65
Cem 1.75 66
Hande 1.62 61
Lale 1.76 64
Arda 1.78 63
Bilgin 1.77 84
Cem 1.69 75
Ozlem 1.75 65
Ali 1.73 75
Haluk 1.71 81All we need is to set the header parameter to TRUE:
hwdata <- read.table("mydata2.txt",header = TRUE)
hwdata Name Height Weight
1 Can 1.70 65
2 Cem 1.75 66
3 Hande 1.62 61
4 Lale 1.76 64
5 Arda 1.78 63
6 Bilgin 1.77 84
7 Cem 1.69 75
8 Ozlem 1.75 65
9 Ali 1.73 75
10 Haluk 1.71 81Now we have another file mydata3.txt whose fields are separated with commas, instead of spaces:
Name,Height,Weight
Can,1.70,65
Cem,1.75,66
Hande,1.62,61
Lale,1.76,64
Arda,1.78,63
Bilgin,1.77,84
Cem,1.69,75
Ozlem,1.75,65
Ali,1.73,75
Haluk,1.71,81To accomodate for that, we set the sep parameter to the separator character, comma.
hwdata <- read.table("mydata3.txt",header = TRUE, sep=",")
hwdata Name Height Weight
1 Can 1.70 65
2 Cem 1.75 66
3 Hande 1.62 61
4 Lale 1.76 64
5 Arda 1.78 63
6 Bilgin 1.77 84
7 Cem 1.69 75
8 Ozlem 1.75 65
9 Ali 1.73 75
10 Haluk 1.71 81Now consider a more complicated data file mydata4.txt, which contains some comments added by the data collector.
Name,Height,Weight
Can,1.70,65
Cem,1.75,66
# Here is a comment
Hande,1.62,61
Lale,1.76,64
Arda,1.78,63
Bilgin,1.77,84 # another comment
Cem,1.69,75
Ozlem,1.75,65
Ali,1.73,75
Haluk,1.71,81The comment character can be set with the comment.char parameter to read.table(). Then, everything on a line starting with # is ignored:
hwdata <- read.table("mydata4.txt",header = TRUE, sep=",", comment.char="#")
hwdata Name Height Weight
1 Can 1.70 65
2 Cem 1.75 66
3 Hande 1.62 61
4 Lale 1.76 64
5 Arda 1.78 63
6 Bilgin 1.77 84
7 Cem 1.69 75
8 Ozlem 1.75 65
9 Ali 1.73 75
10 Haluk 1.71 81Actually, this was a redundant setting, because by default comment.char is already set to "#".
Sometimes the separator character can be used in a text field, such as the space character in the column for names. In such cases, we use quotes to delimit the column’s content, as below (mydata5.txt):
Name Height Weight
"Can Can" 1.70 65
"Cem Cem" 1.75 66
"Hande Hande" 1.62 61
"Lale Lale" 1.76 64
"Arda Arda" 1.78 63
"Bilgin Bilgin" 1.77 84
"Cem Cim" 1.69 75
"Ozlem Ozlem" 1.75 65
"Ali Ali" 1.73 75
"Haluk Haluk" 1.71 81The function read.table() recognizes the single- or double quotes by default.
hwdata <- read.table("mydata5.txt", header=TRUE)
hwdata Name Height Weight
1 Can Can 1.70 65
2 Cem Cem 1.75 66
3 Hande Hande 1.62 61
4 Lale Lale 1.76 64
5 Arda Arda 1.78 63
6 Bilgin Bilgin 1.77 84
7 Cem Cim 1.69 75
8 Ozlem Ozlem 1.75 65
9 Ali Ali 1.73 75
10 Haluk Haluk 1.71 81Other quote characters can be specified using the quote parameter. For example, consider the data file mydata6.txt:
Name Height Weight
%Can Can% 1.70 65
%Cem Cem% 1.75 66
%Hande Hande% 1.62 61
%Lale Lale% 1.76 64
%Arda Arda% 1.78 63
%Bilgin Bilgin% 1.77 84
%Cem Cim% 1.69 75
%Ozlem Ozlem% 1.75 65
%Ali Ali% 1.73 75
%Haluk Haluk% 1.71 81Writing data to a file
Suppose that we process the data file by, e.g., adding some columns.
hwdata <- read.table("mydata6.txt", header=TRUE, quote="%")
hwdata$BMI <- hwdata$Weight / hwdata$Height^2
hwdata$BMI <- round(hwdata$BMI, 2) # round to two decimal places
hwdata Name Height Weight BMI
1 Can Can 1.70 65 22.49
2 Cem Cem 1.75 66 21.55
3 Hande Hande 1.62 61 23.24
4 Lale Lale 1.76 64 20.66
5 Arda Arda 1.78 63 19.88
6 Bilgin Bilgin 1.77 84 26.81
7 Cem Cim 1.69 75 26.26
8 Ozlem Ozlem 1.75 65 21.22
9 Ali Ali 1.73 75 25.06
10 Haluk Haluk 1.71 81 27.70The function write.table() can be used to store a data frame in a file.
write.table(hwdata,"mydata7.txt")This function writes the table together with the row names and column names:
"Name" "Height" "Weight" "BMI"
"1" "Can Can" 1.7 65 22.49
"2" "Cem Cem" 1.75 66 21.55
"3" "Hande Hande" 1.62 61 23.24
"4" "Lale Lale" 1.76 64 20.66
"5" "Arda Arda" 1.78 63 19.88
"6" "Bilgin Bilgin" 1.77 84 26.81
"7" "Cem Cim" 1.69 75 26.26
"8" "Ozlem Ozlem" 1.75 65 21.22
"9" "Ali Ali" 1.73 75 25.06
"10" "Haluk Haluk" 1.71 81 27.7We can omit the row and column names with the following parameter settings.
write.table(hwdata,"mydata7.txt",row.names = FALSE, col.names = FALSE)