根据另一列替换一列中的NA值

Question

Id   authId   sessionId       
139 "56763313.wrpy" "4233a31b52f92c6fe8af4f04f2116657" 
123 "221156400"  "ae04ddacadaa3429ca77dab674a008bf" 
126 "221156400"  "ae04ddacadaa3429ca77dab674a008bf" 
144 "221156400"  "ae04ddacadaa3429ca77dab674a008bf" 
143 "221156400"  "ae04ddacadaa3429ca77dab674a008bf" 
118 NA    "ae04ddacadaa3429ca77dab674a008bf" 
121 NA    "ae04ddacadaa3429ca77dab674a008bf" 
122 NA    "ae04ddacadaa3429ca77dab674a008bf" 
75 "5676614888888" "ca673b5e60a6f70963bf3017e3cb0780" 
276 "56711325.cc79" "f6075188c0f479d7a423744f6c8655b3" 
256 "56711325.cc79" "f6075188c0f479d7a423744f6c8655b3" 
275 "56711325.cc79" "f6075188c0f479d7a423744f6c8655b3" 
152 NA    "f6075188c0f479d7a423744f6c8655b3" 
158 NA    "f6075188c0f479d7a423744f6c8655b3" 
28 "221124184"  "fc71064548bb35d05293bd67d55f1693" 
31 "221124184"  "fc71064548bb35d05293bd67d55f1693" 

我想根据sessionId填补缺失的authId。我试图做到这一点，而不使用循环。例如：

143 "221156400"  "ae04ddacadaa3429ca77dab674a008bf" 
118 "221156400"  "ae04ddacadaa3429ca77dab674a008bf" 

Answer 1

首先创建的authId和sessionId独特的组合一个数据帧。然后找到sessionId任何authId即NA。运用独特的表来查找sessionId的相关authId：

df <- read.table(text="Id   authId   sessionId       
139 56763313.wrpy 4233a31b52f92c6fe8af4f04f2116657 
123 221156400  ae04ddacadaa3429ca77dab674a008bf 
126 221156400  ae04ddacadaa3429ca77dab674a008bf 
144 221156400  ae04ddacadaa3429ca77dab674a008bf 
143 221156400  ae04ddacadaa3429ca77dab674a008bf 
118 NA    ae04ddacadaa3429ca77dab674a008bf 
121 NA    ae04ddacadaa3429ca77dab674a008bf 
122 NA    ae04ddacadaa3429ca77dab674a008bf 
75 5676614888888 ca673b5e60a6f70963bf3017e3cb0780 
276 56711325.cc79 f6075188c0f479d7a423744f6c8655b3 
256 56711325.cc79 f6075188c0f479d7a423744f6c8655b3 
275 56711325.cc79 f6075188c0f479d7a423744f6c8655b3 
152 NA    f6075188c0f479d7a423744f6c8655b3 
158 NA    f6075188c0f479d7a423744f6c8655b3 
28 221124184  fc71064548bb35d05293bd67d55f1693 
31 221124184  fc71064548bb35d05293bd67d55f1693", header=T) 


# find unique combinations of authId and sessionID, but not when authId is NA 
uniques <- unique(df[c("authId", "sessionId")]) 
uniques <- uniques[!is.na(uniques$authId),] 

# replace authID's that are NA with the unique authId associated with the sessionId 
na.authId <- which(is.na(df$authId)) 
na.sessionId <- df$sessionId[na.authId] 
df$authId[na.indices] <- uniques$authId[match(na.sessionId, uniques$sessionId)] 


#  Id  authId      sessionId 
# 1 139 56763313.wrpy 4233a31b52f92c6fe8af4f04f2116657 
# 2 123  221156400 ae04ddacadaa3429ca77dab674a008bf 
# 3 126  221156400 ae04ddacadaa3429ca77dab674a008bf 
# 4 144  221156400 ae04ddacadaa3429ca77dab674a008bf 
# 5 143  221156400 ae04ddacadaa3429ca77dab674a008bf 
# 6 118  221156400 ae04ddacadaa3429ca77dab674a008bf 
# 7 121  221156400 ae04ddacadaa3429ca77dab674a008bf 
# 8 122  221156400 ae04ddacadaa3429ca77dab674a008bf 
# 9 75 5676614888888 ca673b5e60a6f70963bf3017e3cb0780 
# 10 276 56711325.cc79 f6075188c0f479d7a423744f6c8655b3 
# 11 256 56711325.cc79 f6075188c0f479d7a423744f6c8655b3 
# 12 275 56711325.cc79 f6075188c0f479d7a423744f6c8655b3 
# 13 152 56711325.cc79 f6075188c0f479d7a423744f6c8655b3 
# 14 158 56711325.cc79 f6075188c0f479d7a423744f6c8655b3 
# 15 28  221124184 fc71064548bb35d05293bd67d55f1693 
# 16 31  221124184 fc71064548bb35d05293bd67d55f1693