Question

我正在尝试创建与给定日期从API收到的响应相对应的csv文件。

创建一个循环似乎很合乎逻辑，而不是每次我想要一个新日期时都必须编辑代码。我创建了一个称为“日期”的向量，并尝试运行以下代码。

library(httr)
library(jsonlite)
date=c("201801","201802","201803","201804","201805")

for(i in 1:5){
url="https://website.com/api/data"
body=list(fields=list("symbol", "letter", "number"),
          history=date[i])
response=POST(url, body=body, encode="json")
data=content(response)$data       #data is a portion of the response#
write.csv(data[[i]], paste(i, ".csv"))
}

请注意，如果我消除了for循环，而只使用日期向量中的一个元素，则可以获得一个日期所需的输出

url="https://website.com/api/data"
body=list(fields=list("symbol", "letter", "number"),
          history=date[2])
response=POST(url, body=body, encode="json")
data=content(response)$data       #data is a portion of the response#
write.csv(data, '2.csv')

使用for循环会创建一个空响应。关于我要去哪里的任何想法吗？

Answer 1

您可以像这样使它更具可读性：

library(httr)
library(jsonlite)
library(tidyverse)

date <- c("201801","201802","201803","201804","201805")
url <- "https://website.com/api/data"

# Define function
write_files <- function(date, i) {
  body <- list(fields = list("symbol", "letter", "number"),
               history = date)
  response <- POST(url, body = body, encode = "json")
  data <- content(response)$data
  write.csv(data, file = i)
}

# Apply function to each element of date
imap(date, write_files)

（您的示例不可复制，因此未经测试）

Answer 2

如上所述，对于for或lapply之类的迭代运行潜在的开放式进程，请考虑让循环使用Sys.sleep(##)等待进程完成。另外，使用seq_along可以避免 dates 向量的硬编码计数：

dates = c("201801","201802","201803","201804","201805")

for(i in seq_along(dates)){
   url = "https://website.com/api/data"
   body = list(fields = list("symbol", "letter", "number"),
               history = dates[i])
   response = POST(url, body = body, encode = "json")
   data = content(response)$data       # data is a portion of the response

   Sys.sleep(5)     # ADJUST SECONDS AS NEEDED                          
   write.csv(data, paste0(dates[i], ".csv"))
}

或者，使用lapply甚至在将每个csv写入列表以供以后使用之后甚至返回数据：

data_list <- lapply(seq_along(dates), function(i)  {    
     url = "https://website.com/api/data"
     body = list(fields = list("symbol", "letter", "number"),
                 history = dates[i])
     response = POST(url, body = body, encode = "json")
     data = content(response)$data       # data is a portion of the response

     Sys.sleep(5)     # ADJUST SECONDS AS NEEDED                          
     write.csv(data, paste0(dates[i], ".csv"))

     return(data)
})

如何根据来自API的与日期相关的响应编写多个.csv文件

2 个答案: