如何对多个 .csv 文件应用函数？

Question

我有大量使用以下结构命名的 csv 文件： df_climate_a_subset_b（a 范围为 1:12，b 范围为 1:1000）。每个文件如下所示：

names<- c("t1","t2","t3","t4","t5","t6","t7","t8","t9","t10","t11")
value1 <- c(2,3.1,4.5,1,6.5,7.1,8.5,9.11,10.1,4,12.3) 
value2 <- c(2.5,3.1,4.5,2,12,7.1,8.5,10,10.1,17.8,12.3) 
value3 <- c(2,3.1,2,5.1,12,7.1,8.5,9.11,10.1,17.8,12.3) 
value4 <- c(1,3.1,4.5,5.1,12,7.1,8.5,1,10.1,17.8,12.3) 

df_climate_1_subset_1 <- data.frame(names,value1,value2,value3,value4)

我还有一个包含 #b 列的文件。

names<- c("t1","t2","t3","t4","t5","t6","t7","t8","t9","t10","t11")
subset1 <- c(5.5,3,4,1,6,7.1,8.5,9.11,10.1,4,12.3) 
subset2 <- c(2.5,3.1,4.5,2.5,12,7.1,8.5,10,10.1,17.8,12.3) 
subset3 <- c(1,1,1.1,8.5,9,10.1,1,1.5,3,2,4) 
subset4 <- c(1,3.1,4.5,0,12,7.1,0,1,10.1,17.8,12.3) 

subsets_temp <- data.frame(names,subset1,subset2,subset3,subset4)

我很难弄清楚什么是获取所有子集的每个 climte 与来自的相应列之间的相关性的有效方法 subsets_temp 数据帧。

最好的

Answer 1

做一个可重现的例子：

# example of dataset
df_climate_1_subset_1 <- data.frame(
  names  = c("t1","t2","t3","t4","t5","t6","t7","t8","t9","t10","t11"),
  value1 = c(2,3.1,4.5,1,6.5,7.1,8.5,9.11,10.1,4,12.3),
  value2 = c(2.5,3.1,4.5,2,12,7.1,8.5,10,10.1,17.8,12.3),
  value3 = c(2,3.1,2,5.1,12,7.1,8.5,9.11,10.1,17.8,12.3),
  value4 = c(1,3.1,4.5,5.1,12,7.1,8.5,1,10.1,17.8,12.3) 
)
# one df to compare with (100 subsets, not 1000)
subsets_temp <- data.frame(
  names   = c("t1","t2","t3","t4","t5","t6","t7","t8","t9","t10","t11"),
  subset1 = c(5.5,3,4,1,6,7.1,8.5,9.11,10.1,4,12.3),
  subset2 = c(2.5,3.1,4.5,2.5,12,7.1,8.5,10,10.1,17.8,12.3), 
  subset3 = c(1,1,1.1,8.5,9,10.1,1,1.5,3,2,4),
  subset4 = c(1,3.1,4.5,0,12,7.1,0,1,10.1,17.8,12.3) 
)
for (i in 5:100) {
  subset <- paste0("subset", i)
  subsets_temp[[subset]] <- subsets_temp[["subset4"]]
}
# write some files to the disk (in dir tmp/)
for (a in 1:12) {
  for (b in 1:100) {
    filename <- paste("tmp/df_climate", a, "subset", b, sep = "_")
    write.csv(df_climate_1_subset_1, filename, row.names = FALSE)
  }
}

解决方案： 我想所有文件的名称顺序相同。

library(foreach)

res <- foreach(a = 1:12) %:% 
  foreach(b = 1:100, .combine = "cbind") %do% {
    if (b == 1) print(a)
    filename <- paste("tmp/df_climate", a, "subset", b, sep = "_")
    df <- read.csv(filename)
    subset <- paste0("subset", b)
    cor(df[-1], subsets_temp[[subset]])
  }

这为您提供了一个包含 12 个元素的列表（对应于 a 的每个值）。每个元素是一个 4 * 100 的矩阵。

一开始，我以为我们必须使用并行性（这就是我感兴趣的原因），但它似乎超级快，所以没用。