R Language Data remodeling

Data remodeling in the R language is about changing the way data is organized into rows and columns. M ost of the time data processing in the R language is done by using the input data as a data frame. I t is easy to extract data from rows and columns of data frames, but in some cases we need a different format for data frames than the ones we receive for data frames. The R language has many features that split in a data frame, merge and change rows to columns, and vice versa.

Add columns and rows to the data frame

We can use the cbind() function to connect multiple vectors to create data frames. In addition, we can use the rbind() function to combine two data frames.

# Create vector objects.
city <- c("Tampa","Seattle","Hartford","Denver")
state <- c("FL","WA","CT","CO")
zipcode <- c(33602,98104,06161,80294)

# Combine above three vectors into one data frame.
addresses <- cbind(city,state,zipcode)

# Print a header.
cat("# # # # The First data frame
") 

# Print the data frame.
print(addresses)

# Create another data frame with similar columns
new.address <- data.frame(
   city = c("Lowry","Charlotte"),
   state = c("CO","FL"),
   zipcode = c("80230","33949"),
   stringsAsFactors = FALSE
)

# Print a header.
cat("# # # The Second data frame
") 

# Print the data frame.
print(new.address)

# Combine rows form both the data frames.
all.addresses <- rbind(addresses,new.address)

# Print a header.
cat("# # # The combined data frame
") 

# Print the result.
print(all.addresses)

When we execute the code above, it produces the following results -

# # # # The First data frame
     city       state zipcode
[1,] "Tampa"    "FL"  "33602"
[2,] "Seattle"  "WA"  "98104"
[3,] "Hartford" "CT"   "6161" 
[4,] "Denver"   "CO"  "80294"

# # # The Second data frame
       city       state   zipcode
1      Lowry      CO      80230
2      Charlotte  FL      33949

# # # The combined data frame
       city      state zipcode
1      Tampa     FL    33602
2      Seattle   WA    98104
3      Hartford  CT     6161
4      Denver    CO    80294
5      Lowry     CO    80230
6     Charlotte  FL    33949

Merge data frames

We can use the merge() function to combine two data frames. The data frame must have the same column name and be merged on it.

In the following example, we consider the dataset on diabetes in Pima Indian Women in the library name "MASS". W e combine two datasets based on the values of blood pressure ("bp") and body mass index ("bmi"). When you select these two columns for merge, the values of the two variables are combined to form a single data frame by matching records in the two data set.

library(MASS)
merged.Pima <- merge(x = Pima.te, y = Pima.tr,
   by.x = c("bp", "bmi"),
   by.y = c("bp", "bmi")
)
print(merged.Pima)
nrow(merged.Pima)

When we execute the code above, it produces the following results -

   bp  bmi npreg.x glu.x skin.x ped.x age.x type.x npreg.y glu.y skin.y ped.y
1  60 33.8       1   117     23 0.466    27     No       2   125     20 0.088
2  64 29.7       2    75     24 0.370    33     No       2   100     23 0.368
3  64 31.2       5   189     33 0.583    29    Yes       3   158     13 0.295
4  64 33.2       4   117     27 0.230    24     No       1    96     27 0.289
5  66 38.1       3   115     39 0.150    28     No       1   114     36 0.289
6  68 38.5       2   100     25 0.324    26     No       7   129     49 0.439
7  70 27.4       1   116     28 0.204    21     No       0   124     20 0.254
8  70 33.1       4    91     32 0.446    22     No       9   123     44 0.374
9  70 35.4       9   124     33 0.282    34     No       6   134     23 0.542
10 72 25.6       1   157     21 0.123    24     No       4    99     17 0.294
11 72 37.7       5    95     33 0.370    27     No       6   103     32 0.324
12 74 25.9       9   134     33 0.460    81     No       8   126     38 0.162
13 74 25.9       1    95     21 0.673    36     No       8   126     38 0.162
14 78 27.6       5    88     30 0.258    37     No       6   125     31 0.565
15 78 27.6      10   122     31 0.512    45     No       6   125     31 0.565
16 78 39.4       2   112     50 0.175    24     No       4   112     40 0.236
17 88 34.5       1   117     24 0.403    40    Yes       4   127     11 0.598
   age.y type.y
1     31     No
2     21     No
3     24     No
4     21     No
5     21     No
6     43    Yes
7     36    Yes
8     40     No
9     29    Yes
10    28     No
11    55     No
12    39     No
13    39     No
14    49    Yes
15    49    Yes
16    38     No
17    28     No
[1] 17

Sometimes, spreadsheet data is compact enough to give covariables for each topic, followed by all observations for that topic. R 's modeling capabilities need to be observed in a single column. Consider the following sample of data from repeated MRI brain measurements

 Status   Age    V1     V2     V3    V4
      P 23646 45190  50333  55166 56271
     CC 26174 35535  38227  37911 41184
     CC 27723 25691  25712  26144 26398
     CC 27193 30949  29693  29754 30772
     CC 24370 50542  51966  54341 54273
     CC 28359 58591  58803  59435 61292
     CC 25136 45801  45389  47197 47126

There are two covariables and up to four measurements on each topic. The data is exported from Excel .csv file.
We can use the stack to help manipulate this data to give a single response.

zz <- read.csv("mr.csv", strip.white = TRUE)
zzz <- cbind(zz[gl(nrow(zz), 1, 4*nrow(zz)), 1:2], stack(zz[, 3:6]))

The result is:

      Status   Age values ind
X1         P 23646  45190  V1
X2        CC 26174  35535  V1
X3        CC 27723  25691  V1
X4        CC 27193  30949  V1
X5        CC 24370  50542  V1
X6        CC 28359  58591  V1
X7        CC 25136  45801  V1
X11        P 23646  50333  V2
...

The opposite direction of the function unstack may be useful for exporting data.
Another approach is to use function remodeling

> reshape(zz, idvar="id",timevar="var",
  varying=list(c("V1","V2","V3","V4")),direction="long")
    Status   Age var    V1 id
1.1      P 23646   1 45190  1
2.1     CC 26174   1 35535  2
3.1     CC 27723   1 25691  3
4.1     CC 27193   1 30949  4
5.1     CC 24370   1 50542  5
6.1     CC 28359   1 58591  6
7.1     CC 25136   1 45801  7
1.2      P 23646   2 50333  1
2.2     CC 26174   2 38227  2
...

The syntax of reinventing functions is more complex than the stack, but can be used for more than one column of data in a "long" form. If the direction is "wide", the remodeling can also perform the opposite transformation.