r - 在现有数据框中添加向量作为新列

Question

示例数据

df <- data.frame(location = rep(1:1000, each = 36), 
                 year = rep(1980:2015,times = 1000),
                 mu = runif(min =  36.5, max = 43.2, 1000*36),
                 lambda = runif(min =  4.5, max = 4.8, 1000*36))

该数据由 1000 个地点和 36 年组成，有两个变量mu 和 lambda

对于每个位置 X 年组合，我有一个函数，需要lambda 和 mu 的值并生成大小为 12 的向量。示例:

library(grofit)    
dat <- df[df$location == 1 & df$year == 1980,]  
y <- round(gompertz(1:12,100,dat$mu,dat$lambda), digits = 2)
y

  [1]  0.00  0.00  0.00  0.72 18.60 56.37 82.26 93.56 97.76 99.23
  [11] 99.74 99.91

如果我想将 y 作为列添加到 dat

  new.col <- 5:16 
  dat[new.col] <- y
  dat

  location year       mu   lambda V5 V6 V7   V8   V9   V10   V11
     1     1980 39.60263 4.554095  0  0  0 0.72 18.6 56.37 82.26
  V12   V13   V14   V15   V16
  1 93.56 97.76 99.23 99.74 99.91

如您所见，我已将 y 作为 dat 中的 V5 至 V16 列附加。我想对 df 中的所有位置和年份组合重复此操作。我希望这一点很清楚。

df %>% group_by(location year) %>% mutate(?? how to I add new columns for y??)

Answer 1

您可以使用lapply():

library(grofit)    
df2 <- do.call(rbind, lapply(1:nrow(df), 
                             function(x) round(gompertz(1:12, 100, df[x, 3], df[x, 4]), 
                                               digits = 2)))
df3 <- cbind(df, df2)

结果:

> head(df3)
  location year       mu   lambda 1 2 3    4     5     6     7     8     9    10    11    12
1        1 1980 43.04565 4.536717 0 0 0 0.61 20.58 61.23 85.88 95.39 98.54 99.55 99.86 99.96
2        1 1981 39.00524 4.505235 0 0 0 0.96 20.02 57.28 82.45 93.53 97.71 99.20 99.72 99.90
3        1 1982 41.60206 4.619627 0 0 0 0.42 17.07 56.52 83.18 94.23 98.10 99.38 99.80 99.94
4        1 1983 42.01069 4.689058 0 0 0 0.26 14.87 54.43 82.35 93.99 98.04 99.37 99.80 99.94
5        1 1984 40.34275 4.692595 0 0 0 0.30 14.36 52.30 80.54 93.03 97.61 99.20 99.73 99.91
6        1 1985 41.13246 4.641404 0 0 0 0.38 16.20 55.15 82.32 93.84 97.94 99.32 99.78 99.93

数据:

set.seed(47)  # for sake of reproducibility
df <- data.frame(location = rep(1:1000, each = 36), 
                 year = rep(1980:2015, times = 1000),
                 mu = runif(min =  36.5, max = 43.2, 1000 * 36),
                 lambda = runif(min =  4.5, max = 4.8, 1000 * 36))