Create DataFrame Row by Row in R

In this article, we will discuss how to create dataframe row by row in R Programming Language.

Method 1: Using for loop and indexing methods

An empty data frame in R language can be created using the data.frame() method in R. For better clarity, the data types of the columns can be defined during the declaration. Each row of the data frame is a vector consisting of values belonging to different columns. The i^th row in the data frame can then be assigned to this vector. 

Syntax:

dataframe[ i, ] <- vec

The rows are appended to the end of the data frame. The addition of rows can also be done in a for loop, with the loop iterations equivalent to the number of rows to add.  The length of the vector appended should be equivalent to the number of columns in the data frame. 

Example:

R

# declaring an empty data frame
data_frame = data.frame(
  col1 = numeric(), col2 = character(),stringsAsFactors = FALSE)
print ("Original dataframe")
print (data_frame)

# appending rows to the data frame
for(i in 1:3) {

  # creating a vector to append to
  # data frame
  vec <- c(i+1, LETTERS[i])

  # assigning this vector to ith row
  data_frame[i, ] <- vec
}

print ("Modified dataframe")
print (data_frame)

Output

[1] "Original dataframe"
[1] col1 col2
<0 rows> (or 0-length row.names)
[1] "Modified dataframe"
 col1 col2
1    2    A
2    3    B
3    4    C

Method 2 : Using manual insertion

The rows in an empty data frame can also be inserted using the row index reference. After each iteration, a new row is appended to the data frame. Modifications are made to the original data frame.

Example:

R

# declaring an empty data frame
data_frame = data.frame(col1 = numeric(),
                        col2 = character(),
                        stringsAsFactors = FALSE)

print ("Original dataframe")
print (data_frame)

# appending rows to the data frame
data_frame[1,] <- c(1,"firstrow")

print ("Dataframe : Iteration 1")
print (data_frame)

# appending rows to the data frame
data_frame[2,] <- c(2,"secondrow")
print ("Dataframe : Iteration 2")
print (data_frame)

data_frame[3,] <- c(3,"thirdrow")
print ("Dataframe : Iteration 3")
print (data_frame)

Output

[1] "Original dataframe"
[1] col1 col2
<0 rows> (or 0-length row.names)
[1] "Dataframe : Iteration 1"
 col1     col2
1    1 firstrow
[1] "Dataframe : Iteration 2"
 col1      col2
1    1  firstrow
2    2 secondrow
[1] "Dataframe : Iteration 3"
 col1      col2
1    1  firstrow
2    2 secondrow
3    3  thirdrow

The current number of rows of the data frame can be captured by the nrow(dataframe) method. An individual row can be added to the data frame at the nrow(df)+1 index. 

Syntax:

df[ nrow(df)+1, ] <- vec

Example:

R

# declaring an empty data frame
data_frame = data.frame(col1 = c(2:3),
                        col2 = letters[1:2],
                        stringsAsFactors = FALSE)

print ("Original dataframe")
print (data_frame)

# appending rows at the end
vec <- c(8,"n")
data_frame[nrow(data_frame)+1,] <- vec

print ("Modified dataframe")
print (data_frame)

Output

[1] "Original dataframe"
 col1 col2
1    2    a
2    3    b
[1] "Modified dataframe"
 col1 col2
1    2    a
2    3    b
3    8    n

Method 3 : Using rbind() method

rbind() method is used to bind vectors or data frame objects together to form a larger object. Initially, we can create an empty data frame and then define a new row using a vector, and bind the declared row to the data frame using the rbind() method. The new row is appended at the end. The data types of the row should be compatible with the data types declared for the original data frame. This can be enclosed within a loop to bind rows during each loop iteration. 

Syntax:

rbind(df , vec)

Example:

R

# declaring an empty data frame
data_frame = data.frame(col1 = numeric(),
                        col2 = character(),
                        stringsAsFactors = FALSE)

print ("Original dataframe")
print (data_frame)

# appending rows at the end
vec <- c(8,"n")

# binding row at the end of the data frame
data_frame <- rbind(data_frame,vec,stringsAsFactors = FALSE)

print ("Modified dataframe")
print (data_frame)

Output

[1] "Original dataframe"
[1] col1 col2
<0 rows> (or 0-length row.names)
[1] "Modified dataframe"
  X.8. X.n.
1    8    n

A data frame can also be created row by row, using repeated rbind() operations on the data frame in the for loop, with number of iterations equivalent to the number of rows to insert.

Example:

R

# declaring an empty data frame
data_frame = data.frame(col1 = numeric(),
                        col2 = character(),
                        stringsAsFactors = FALSE)

print ("Original dataframe")
print (data_frame)

# appending rows at the end
for(i in 1:3) {

  # creating a vector to append
  # to data frame
  vec <- c(i, LETTERS[i])

  # assigning this vector to ith row
  data_frame <- rbind(data_frame,vec,stringsAsFactors=FALSE)
}

print ("Modified dataframe")
print (data_frame)

Output

[1] "Original dataframe"
[1] col1 col2
<0 rows> (or 0-length row.names)
[1] "Modified dataframe"
  X.1. X.A.
1    1    A
2    2    B
3    3    C