R Notebook

So we want to add person-level summaries of a variable to the right of a data-frame?

1. Create some data

library(dplyr)

## 
## Attaching package: 'dplyr'

## The following objects are masked from 'package:stats':
## 
##     filter, lag

## The following objects are masked from 'package:base':
## 
##     intersect, setdiff, setequal, union

suppressMessages(library(data.table))

d <- data.frame(ID = rep(seq(1,5,1), each = 5),
                X  = rnorm(25))
head(d) %>% knitr::kable()

ID	X
1	-0.0384933
1	-0.2161823
1	-0.2506520
1	-0.2967449
1	-0.6339499
2	0.5708115

2. The dplyr way

d %>%
  group_by(ID) %>% 
  mutate(mean_of_X = mean(X)) %>%
  head(.) %>% knitr::kable()

ID	X	mean_of_X
1	-0.0384933	-0.2872045
1	-0.2161823	-0.2872045
1	-0.2506520	-0.2872045
1	-0.2967449	-0.2872045
1	-0.6339499	-0.2872045
2	0.5708115	0.1678269

3. The data.table way

DT <- as.data.table(d)
DT[, mean_of_X := mean(X), by = ID]
head(DT) %>% knitr::kable()

ID	X	mean_of_X
1	-0.0384933	-0.2872045
1	-0.2161823	-0.2872045
1	-0.2506520	-0.2872045
1	-0.2967449	-0.2872045
1	-0.6339499	-0.2872045
2	0.5708115	0.1678269

4. The base R way

d <- merge(d, aggregate(d$X, by=list(d$ID), FUN=mean), by.x = 'ID', by.y = 'Group.1')
head(d) %>% knitr::kable()

ID	X	x
1	-0.0384933	-0.2872045
1	-0.2161823	-0.2872045
1	-0.2506520	-0.2872045
1	-0.2967449	-0.2872045
1	-0.6339499	-0.2872045
2	0.5708115	0.1678269