Calculate Group Mean While Excluding Current Observation Using Dplyr

Calculate group mean while excluding current observation using dplyr

No need to define a custom function, instead we could simply sum all elements of the group, subtract the current value, and divide by number of elements per group minus 1.

df %>% group_by(grouping) %>%
        mutate(special_mean = (sum(value) - value)/(n()-1))
#   grouping value special_mean
#      (chr) (int)        (dbl)
#1         A     1          8.5
#2         A     6          6.0
#3         A    11          3.5
#4         B     2          9.5
#5         B     7          7.0

Calculate group variance while excluding current observation

Try the following :

library(dplyr)

DF %>% 
  group_by(School)%>%
  mutate(Var_grade = purrr::map_dbl(row_number(), ~var(grade[-.x])))

#  School grade Var_grade
#   <int> <dbl>     <dbl>
#1      1    90     112. 
#2      1    80      12.5
#3      1    95      50  
#4      2   100     108. 
#5      2    65     225  
#6      2    70     308. 
#7      2    85     358. 

---

In base you can use ave with sapply :

DF$Var_grade <- with(DF, ave(grade, School, FUN = function(x) 
                     sapply(seq_along(x), function(i) var(x[-i]))))

data

DF <- data.frame(School = rep(1:2, c(3, 4)), 
                 grade = c(90, 80, 95, 100, 65, 70, 85))

Exclude current observation from computation in dplyr pipe

For a general case to remove current observation and perform calculation, you could use map_dbl

library(dplyr)
library(purrr)
da %>%
  group_by(ice_id) %>%
  mutate(mean_price = mean(price),
         mean_price_without = map_dbl(day, ~mean(price[-.x])))
         #Or
         #mean_price_without = map_dbl(day, ~mean(price[day != .x])))
         #mean_price_without = map_dbl(row_number(), ~mean(price[-.x])))


#  ice_id   day price mean_price mean_price_without
#   <dbl> <dbl> <dbl>      <dbl>              <dbl>
#1      1     1  1.6       1.77               1.85 
#2      1     2  1.9       1.77               1.7  
#3      1     3  1.8       1.77               1.75 
#4      2     1  2.1       2.15               2.17 
#5      2     2  2.05      2.15               2.2  
#6      2     3  2.3       2.15               2.08 
#7      3     1  0.5       0.417              0.375
#8      3     2  0.4       0.417              0.425
#9      3     3  0.35      0.417              0.45 

Get group mean with multiple grouping variables and excluding own group value

library(dplyr)

df %>%
  group_by(state, year) %>%
  mutate(q = (sum(value) - value) / (n()-1))

#> # A tibble: 12 x 5
#> # Groups:   state, year [4]
#>    state county  year value     q
#>    <chr> <chr>  <int> <int> <dbl>
#>  1 AL    a       2011    68  30.5
#>  2 AL    a       2012    63  42  
#>  3 AL    b       2011    53  38  
#>  4 AL    b       2012    56  45.5
#>  5 AL    c       2011     8  60.5
#>  6 AL    c       2012    28  59.5
#>  7 CA    d       2011     7  40  
#>  8 CA    d       2012    69  41  
#>  9 CA    e       2011    39  24  
#> 10 CA    e       2012    79  36  
#> 11 CA    f       2011    41  23  
#> 12 CA    f       2012     3  74

Data:

#data_frame is deprecate!
df <- tibble(
  state = rep(c("AL", "CA"), each = 6),
  county = rep(letters[1:6], each = 2),
  year = rep(c(2011:2012), 6),
  value = sample.int(100, 12)
)

Taking group means, excluding the observation itself (and dealing with NA's)

IIUC this should do what you're looking for:

DF[, mean_value := (sum(value, na.rm=TRUE)-value)/(sum(!is.na(value))-!is.na(value)),  
     by=c("iso", "year")]

    A B D value iso year mean_value
 1: 0 1 1    NA ECU 2009         NA
 2: 1 0 2     1 ECU 2009        2.0
 3: 1 0 1     2 ECU 2009        1.0
 4: 0 0 3     1 BRA 2011        0.5
 5: 1 0 4     0 BRA 2011        1.0
 6: 0 0 3     1 BRA 2011        0.5
 7: 0 1 7    NA ECU 2008         NA
 8: 1 0 1     1 ECU 2008        1.0
 9: 1 0 1     1 ECU 2008        1.0
10: 0 0 3     2 BRA 2012        2.0
11: 0 0 3     2 BRA 2012        2.0
12: 1 0 4    NA BRA 2012         NA

Note: you may want to additionally consider edge cases like a group of size 1 with NA value which would lead to division by zero

Compute mean excluding current value

I am not sure if your calculation is correct for group 1 but you can do -

library(data.table)

setDT(df)[, avg2 := (sum(b) - b)/(.N -1), a]
df

#   a b avg avg2
#1: 1 7   3  1.0
#2: 1 0   3  4.5
#3: 1 2   3  3.5
#4: 2 1   2  3.0
#5: 2 3   2  1.0

Calculate standard deviation by group excluding current observation in R

An option is to use dplyr and mapply. mapply runs for every row (of group) and sd calculation excludes the current row.

library(dplyr)

df %>% group_by(country) %>%
  mutate(Sp_SD = mapply(function(x)sd(weight[-x]), 1:n()))


# # A tibble: 6 x 3
# # Groups: country [2]
# country weight Sp_SD
# <fctr>   <dbl> <dbl>
# 1 A         10.0 0.707
# 2 A         11.0 1.41 
# 3 A         12.0 0.707
# 4 B         20.0 3.54 
# 5 B         25.0 7.07 
# 6 B         30.0 3.54 

dplyr mutate: Excluding observations similar to the current one

You can likely do this more succinctly, but this will get you the result.

You essentially create a column which contains the total observations and sum of records for the whole data.frame. Then you group by the X column and repeat the process, by taking the difference you can calculate your mean.

data

df <- data.frame(X = c("A", "A", "B", "B", "C", "C"),
                 Y = c(1:6))

---

solution

library(tidyverse)
df %>%
  mutate(total_sum = sum(Y),
         total_obs = n()) %>%
  group_by(X) %>%
  mutate(group_sum = sum(Y),
         group_obs = n()) %>%
  ungroup() %>%
  mutate(other_group_sum = total_sum - group_sum,
         other_group_obs = total_obs - group_obs,
         other_mean = other_group_sum/other_group_obs) %>%
  select(X, Y, other_mean)

---

result

# A tibble: 6 x 3
  X         Y other_mean
  <fct> <int>      <dbl>
1 A         1       4.50
2 A         2       4.50
3 B         3       3.50
4 B         4       3.50
5 C         5       2.50
6 C         6       2.50

---

Related Topics