Reading text within the source file into a dataframe
With the
text= parameter set, it's possible to read text that is stored
within the source file itself into a
data frame.
Note how read.table is clever enough to recognize the different data types of the data read.
df <- read.table(
text='name val_1 val_2
foo 42 18.8
bar 123456789 -28.2
baz 0 99.9',
header = TRUE)
class(df) # data.frame
mode(df) # list
typeof(df) # list
df
#
# name val_1 val_2
# 1 foo 42 18.8
# 2 bar 123456789 -28.2
# 3 baz 0 99.9
class (df$name ) # factor
class (df$val_1) # integer
class (df$val_2) # numeric
mode (df$name ) # numeric
mode (df$val_1) # numeric
mode (df$val_2) # numeric
typeof(df$name ) # integer
typeof(df$val_1) # integer
typeof(df$val_2) # double
Quoted strings
If the text to be read contains spaces, it can be quoted:
df <- read.table(
text="name val_1
'item one' 1.1
'item two' 2.22
'item three' 3.333",
header = TRUE)
df
#
# name val_1
# 1 item one 1.100
# 2 item two 2.220
# 3 item three 3.333
na.strings
The
na.strings= argument specifies a string constant that should be interpreted as
NA when reading the data.
df <- read.table(
header = TRUE ,
na.strings ='null',
text ='id val
1 foo
2 null
3 bar
4 baz'
)
df
#
# id val
# 1 1 foo
# 2 2 <NA>
# 3 3 bar
# 4 4 baz
Reading data from the internet
#
# https://stackoverflow.com/a/1296434/180275
#
url <- paste0(
'https://random.org/integers/',
'?num=100' , # Download 100 numbers
'&base=10' , # with base 10
'&min=0' , # in the range of 0 …
'&max=100' , # 100
'&col=2' , # Two columns
'&format=plain' , # text mode
'&rnd=new'
);
tab <- read.table(url);