I have these scores per participant and I want to sum them all up, but there's one condition: participants cannot make 4 errors in a row, so if there's four consecutive zeros, the rest of the scores should not count. Is there a simple way of going about it? In this example, pt4 would be the problematic one and the last two scores should be removed from the count.
library(dplyr)
pt1 <- rep(1, 44)
pt2 <- c(rep(1,35), rep(0, 9))
pt3 <- c(rep(1,37), rep(0, 7))
pt4 <- c(rep(1,38), rep(0, 4), rep(1,2))
df <- data.frame(item = rep(1:44, 4), pt = rep(LETTERS[1:4], each = 44), acc = c(pt1, pt2, pt3, pt4))
df %>% group_by(pt) %>%
dplyr::summarise(
total = sum(as.numeric(acc)))
CodePudding user response：
This can be resolved using run-length encoding (rle). To really make sure this works, I'll add some less-than-4-long 0s in the data.
df$acc[4:6] <- 0
A quick function:
func <- function(z) {
r <- rle(z)
r$values <- !cumsum(r$lengths >= 4 & !r$values) > 0
inverse.rle(r)
}
And using it.
base R
df[ave(df$acc != 0, df$pt, FUN = func),]
# item pt acc
# 1 1 A 1
# 2 2 A 1
# 3 3 A 1
# 4 4 A 0
# 5 5 A 0
# 6 6 A 0
# 7 7 A 1
# 8 8 A 1
# 9 9 A 1
# 10 10 A 1
# 11 11 A 1
# 12 12 A 1
# 13 13 A 1
# 14 14 A 1
# 15 15 A 1
# 16 16 A 1
# 17 17 A 1
# 18 18 A 1
# 19 19 A 1
# 20 20 A 1
# 21 21 A 1
# 22 22 A 1
# 23 23 A 1
# 24 24 A 1
# 25 25 A 1
# 26 26 A 1
# 27 27 A 1
# 28 28 A 1
# 29 29 A 1
# 30 30 A 1
# 31 31 A 1
# 32 32 A 1
# 33 33 A 1
# 34 34 A 1
# 35 35 A 1
# 36 36 A 1
# 37 37 A 1
# 38 38 A 1
# 39 39 A 1
# 40 40 A 1
# 41 41 A 1
# 42 42 A 1
# 43 43 A 1
# 44 44 A 1
# 45 1 B 1
# 46 2 B 1
# 47 3 B 1
# 48 4 B 1
# 49 5 B 1
# 50 6 B 1
# 51 7 B 1
# 52 8 B 1
# 53 9 B 1
# 54 10 B 1
# 55 11 B 1
# 56 12 B 1
# 57 13 B 1
# 58 14 B 1
# 59 15 B 1
# 60 16 B 1
# 61 17 B 1
# 62 18 B 1
# 63 19 B 1
# 64 20 B 1
# 65 21 B 1
# 66 22 B 1
# 67 23 B 1
# 68 24 B 1
# 69 25 B 1
# 70 26 B 1
# 71 27 B 1
# 72 28 B 1
# 73 29 B 1
# 74 30 B 1
# 75 31 B 1
# 76 32 B 1
# 77 33 B 1
# 78 34 B 1
# 79 35 B 1
# 89 1 C 1
# 90 2 C 1
# 91 3 C 1
# 92 4 C 1
# 93 5 C 1
# 94 6 C 1
# 95 7 C 1
# 96 8 C 1
# 97 9 C 1
# 98 10 C 1
# 99 11 C 1
# 100 12 C 1
# 101 13 C 1
# 102 14 C 1
# 103 15 C 1
# 104 16 C 1
# 105 17 C 1
# 106 18 C 1
# 107 19 C 1
# 108 20 C 1
# 109 21 C 1
# 110 22 C 1
# 111 23 C 1
# 112 24 C 1
# 113 25 C 1
# 114 26 C 1
# 115 27 C 1
# 116 28 C 1
# 117 29 C 1
# 118 30 C 1
# 119 31 C 1
# 120 32 C 1
# 121 33 C 1
# 122 34 C 1
# 123 35 C 1
# 124 36 C 1
# 125 37 C 1
# 133 1 D 1
# 134 2 D 1
# 135 3 D 1
# 136 4 D 1
# 137 5 D 1
# 138 6 D 1
# 139 7 D 1
# 140 8 D 1
# 141 9 D 1
# 142 10 D 1
# 143 11 D 1
# 144 12 D 1
# 145 13 D 1
# 146 14 D 1
# 147 15 D 1
# 148 16 D 1
# 149 17 D 1
# 150 18 D 1
# 151 19 D 1
# 152 20 D 1
# 153 21 D 1
# 154 22 D 1
# 155 23 D 1
# 156 24 D 1
# 157 25 D 1
# 158 26 D 1
# 159 27 D 1
# 160 28 D 1
# 161 29 D 1
# 162 30 D 1
# 163 31 D 1
# 164 32 D 1
# 165 33 D 1
# 166 34 D 1
# 167 35 D 1
# 168 36 D 1
# 169 37 D 1
# 170 38 D 1
dplyr
library(dplyr)
df %>%
group_by(pt) %>%
filter(func(acc != 0)) %>%
ungroup()
# # A tibble: 154 x 3
# item pt acc
# <int> <chr> <dbl>
# 1 1 A 1
# 2 2 A 1
# 3 3 A 1
# 4 4 A 0
# 5 5 A 0
# 6 6 A 0
# 7 7 A 1
# 8 8 A 1
# 9 9 A 1
# 10 10 A 1
# # ... with 144 more rows