Question

我一直在阅读类似的帖子，我不能让任何解决方案适用于我的情况（可能是因为我是R的新手）。

我有一个包含多个参数的长数据集，其中一个是日期，另一个数据框的日期间隔对应于特定值。我试图制作一个可重复的例子：

df = data.frame(date=c("2017/08/01 19:00:00","2017/08/01 19:10:00","2017/08/01 19:20:00","2017/08/01 19:30:00",
                   "2017/08/01 19:40:00","2017/08/01 19:50:00","2017/08/01 20:00:00","2017/08/01 20:10:00"),
            factor1=c(10,15,11,13,14,12,16,15))

df2 = data.frame(start=c("2017/08/01 19:00:00","2017/08/01 19:40:00"),
             end=c("2017/08/01 19:15:00","2017/08/01 20:05:00"), factor2=c("A","B"))

df$date <- as.POSIXct(df$date) 
df2$start <- as.POSIXct(df2$start)
df2$end  <- as.POSIXct(df2$end)

我想要的结果是这样的：

result = data.frame(date=c("2017/08/01 19:00:00","2017/08/01 19:10:00","2017/08/01 19:20:00","2017/08/01 19:30:00",
                   "2017/08/01 19:40:00","2017/08/01 19:50:00","2017/08/01 20:00:00","2017/08/01 20:10:00"),
            factor1=c(10,15,11,13,14,12,16,15),factor2=c("A","A","NA","NA","B","B","B","NA"))

我尝试了ifelse：

ifelse(df$date >= df2$start & df$date <= df2$end,df2$factor2,"NA")

但无法使其发挥作用。

有什么建议吗？

Answer 1

这也适用于您的样本数据：

result <- df
result$factor2 <- NA
for (i in seq_along(df$date)){
  p <- ifelse(length(grep("TRUE", (df$date[i] >= df2$start & df$date[i] <= df2$end)))!=0,
              grep("TRUE", (df$date[i] >= df2$start & df$date[i] <= df2$end)),
              NA)
  result$factor2[i] <- ifelse(!is.na(p),
                          as.character(df2$factor2[p]),
                          "NA")
  }
print(result)
#                 date factor1 factor2
#1 2017-08-01 19:00:00      10       A
#2 2017-08-01 19:10:00      15       A
#3 2017-08-01 19:20:00      11      NA
#4 2017-08-01 19:30:00      13      NA
#5 2017-08-01 19:40:00      14       B
#6 2017-08-01 19:50:00      12       B
#7 2017-08-01 20:00:00      16       B
#8 2017-08-01 20:10:00      15      NA

Answer 2

我尝试在//Get all the names of the contacts and their phone number ContentResolver cr = ContentResolver; ICursor cur = cr.Query(ContactsContract.Contacts.ContentUri,null, null, null, null); if (cur.Count > 0) { while (cur.MoveToNext()) { String id = cur.GetString(cur.GetColumnIndex(ContactsContract.Contacts.InterfaceConsts.Id)); String name = cur.GetString(cur.GetColumnIndex(ContactsContract.Contacts.InterfaceConsts.DisplayName)); if (cur.GetInt(cur.GetColumnIndex(ContactsContract.Contacts.InterfaceConsts.HasPhoneNumber)) > 0) { ICursor pCur = cr.Query(ContactsContract.CommonDataKinds.Phone.ContentUri, null, ContactsContract.CommonDataKinds.Phone.InterfaceConsts.ContactId + " = ?", new String[] { id }, null); while (pCur.MoveToNext()) { String phoneNo = pCur.GetString(pCur.GetColumnIndex(ContactsContract.CommonDataKinds.Phone.Number)); System.Diagnostics.Debug.WriteLine("Name: " + name + ", Phone No: " + phoneNo); } pCur.Close(); } } }中使用inner join，但似乎有效：

sqldf

（请注意，我也尝试使用library(sqldf) df3 = sqldf("select df.*, df2.factor2 from df inner join df2 where df.date >= df2.start and df.date <= df2.end") result = merge(df, df3, by = "date", all.x = TRUE)代替left outer join，但这给了我与inner join相同的结果......它必须是{{1}中的侥幸}）

将数据框中的日期与其他数据框中的两个日期进行比较

2 个答案: