Binding

“In God we trust, all others bring data.” — W Edwards Deming

Before you start: table selection; function calls and returned values.
Setup: No input files are needed. Build the example tables on this page before modifying them.
Your check: Distinguish adding observations from matching attributes by identifiers. Practise in the unit summary.

When thinking about combining two data frames one has to distinguish between merging them by the values given in a specific column or consecutively putting them together row by row.

Unit 03 used binding to construct objects. Here, first decide whether you are adding observations or matching attributes, and check column names, types and units. Matching belongs in the merge lesson.

For the following examples, we will use these two data frames.

x <- c(1, 2, 3, 4, 5)
y <- c(1.4, 2.5, 3.6, 4.0, 5.5)
z <- c("A", "B", "C", "D", "E")

df1 <- data.frame(X = x, Y = y, Z = z)
df1

##   X   Y Z
## 1 1 1.4 A
## 2 2 2.5 B
## 3 3 3.6 C
## 4 4 4.0 D
## 5 5 5.5 E

y <- c(14, 25, 36, 40, 55)
x <- c(100, 200, 300, 400, 500)
z <- c("A", "C", "E", "D", "B")
df2 <- data.frame(X = x, Y = y, Z = z)
df2

##     X  Y Z
## 1 100 14 A
## 2 200 25 C
## 3 300 36 E
## 4 400 40 D
## 5 500 55 B

To combine two data frames, use the rbind function provided that both data frames have the same column names and the same number of columns.

dfr <- rbind(df1, df2)
dfr

##      X    Y Z
## 1    1  1.4 A
## 2    2  2.5 B
## 3    3  3.6 C
## 4    4  4.0 D
## 5    5  5.5 E
## 6  100 14.0 A
## 7  200 25.0 C
## 8  300 36.0 E
## 9  400 40.0 D
## 10 500 55.0 B

Note that it is not necessary that the columns are in the same order since the concatenation is done by the column names, not the column sequence:

df3 <- df2[,c(1,3,2)]
df3

##     X Z  Y
## 1 100 A 14
## 2 200 C 25
## 3 300 E 36
## 4 400 D 40
## 5 500 B 55

dfr <- rbind(df1, df3)
dfr

##      X    Y Z
## 1    1  1.4 A
## 2    2  2.5 B
## 3    3  3.6 C
## 4    4  4.0 D
## 5    5  5.5 E
## 6  100 14.0 A
## 7  200 25.0 C
## 8  300 36.0 E
## 9  400 40.0 D
## 10 500 55.0 B

Check row correspondence before adding columns

The Z labels in df1 and df2 have different row orders. Before running the following call, predict which second-table value it would attach to the B row in the first table:

by_position <- cbind(df1, second_X = df2$X)
by_position

It attaches 200, whose Z label in df2 is C. The matching B value is 500. cbind() aligns positions; it does not match Z. If Z identifies the same entities in both tables, use merge(df1, df2, by = "Z") and check key uniqueness. If you are adding observations with compatible columns instead, use rbind(). Similar names alone do not establish that units or observation meanings agree.