在 R 中对多列使用聚集函数时出错 "Attribute not identical"
Error "Attribute not identical" in using gather function for the multi column in R
我正在使用以下数据框。
df2<-final.data%>% gather(Hospital,Attendance,contains("Attendance"))
df2 %>% spread(Hospital, Attendance)
> dput(final.data[0:2,])
structure(list(RoyalPerth.Attendance = c(235, 209), RoyalPerth.Admissions = c(99,
97), RoyalPerth.Tri1 = c("8", "N/A"), RoyalPerth.Tri2 = c(33,
41), RoyalPerth.Tri3 = c(89, 73), RoyalPerth.Tri4 = c(85, 80),
RoyalPert
h.Tri5 = c("20", "14"), Fremantle.Attendance = c(155,
145), Fremantle.Admissions = c(70, 56), Fremantle.Tri1 = c("N/A",
"N/A"), Fremantle.Tri2 = c(25, 22), Fremantle.Tri3 = c(67,
51), Fremantle.Tri4 = c(54, 47), Fremantle.Tri5 = c(9, 24
), PrincessMargaret.Attendance = c(252, 219), PrincessMargaret.Admissions = c(59,
47), PrincessMargaret.Tri1 = c("N/A", "N/A"), PrincessMargaret.Tri2 = c("13",
"14"), PrincessMargaret.Tri3 = c(75, 61), PrincessMargaret.Tri4 = c(159,
139), PrincessMargaret.Tri5 = c("4", "4"), KingEdward.Attendance = c(52,
43), KingEdward.Admissions = c("6", "7"), KingEdward.Tri1 = c("N/A",
"N/A"), KingEdward.Tri2 = c("N/A", "N/A"), KingEdward.Tri3 = c("7",
"N/A"), KingEdward.Tri4 = c(20, 25), KingEdward.Tri5 = c("25",
"17"), SirCharles.Attendance = c(209, 184), SirCharles.Admissions = c(109,
112), SirCharles.Tri1 = c("N/A", "N/A"), SirCharles.Tri2 = c(42,
43), SirCharles.Tri3 = c(108, 73), SirCharles.Tri4 = c(47,
61), SirCharles.Tri5 = c("11", "5"), Armadale.Attendance = c(166,
175), Armadale.Admissions = c(19, 25), Armadale.Tri1 = c("N/A",
"N/A"), Armadale.Tri2 = c(16, 26), Armadale.Tri3 = c(62,
73), Armadale.Tri4 = c(79, 55), Armadale.Tri5 = c("9", "19"
), Swan.Attendance = c(133, 129), Swan.Admissions = c(17,
25), Swan.Tri1 = c("N/A", "N/A"), Swan.Tri2 = c(29, 25),
Swan.Tri3 = c(59, 57), Swan.Tri4 = c(42, 43), Swan.Tri5 = c("N/A",
"4"), Rockingham.Attendance = c(155, 145), Rockingham.Admissions = c("10",
"24"), Rockingham.Tri1 = c("N/A", "N/A"), Rockingham.Tri2 = c(12,
26), Rockingham.Tri3 = c(51, 45), Rockingham.Tri4 = c(81,
65), Rockingham.Tri5 = c("11", "8"), Joondalup.Attendance = c(267,
241), Joondalup.Admissions = c(73, 81), Joondalup.Tri1 = c("N/A",
"N/A"), Joondalup.Tri2 = c(27, 23), Joondalup.Tri3 = c(75,
78), Joondalup.Tri4 = c(151, 133), Joondalup.Tri5 = c("12",
"7")), row.names = 1:2, class = "data.frame")
错误：
警告信息：
测量变量之间的属性不相同；
他们将被删除
我试过以下方法：
hospital.dataset<-gather(hospital,triage,sum,Tri1:Tri5) 收集分类
在数据集上使用 cbind 之后。
我想使用 gather 将其转换为长数据集。
dput(hospital.dataset[1:2,])
structure(list(Date = structure(c(-714598, -714597), class = "Date"), [enter image description here][1]
Attendance = c(235, 209), Admissions = c(99, 97), Hospital = structure(c(1L,
1L), .Label = c("RoyalPerth Hospital", "Fremantle Hospital",
"Princess Margaret Hospital", "KingEdward Hospital", "SirCharles Hospital",
"Armadale Hospital", "Swan Hospital", "Rockingham Hospital",
"Joondalup Hospital"), class = "factor"), triage = c("Tri1",
"Tri1"), sum = c(8, 0)), row.names = 1:2, class = "data.frame")
像这样。
提前致谢。
Expected Dataframe
注：这个方案感觉很费力。所以请考虑可能有更优雅的方法可用。
此数据的一个问题是您想要“宽”值（Attendance、Admissions）与您想要“长”值（Tri1、Tri1、 Tri2，等等）。
这个解决方案在整个数据框上使用pivot_longer（注意：pivot_longer是新的gather语法），然后separate拉出医院名称来自特定数据字段。
然后它分成两个数据帧，将pivot_wider应用于Attendance/Admissions列，然后重新加入。
library(tidyverse)
final_data_long <- final.data.raw %>%
mutate_all(as.character) %>%
mutate(row_n = row_number()) %>%
pivot_longer(-row_n, names_to = "field", values_to = "value") %>%
separate(field, into = c("hospital", "category"))
attend_admit <- final_data_long %>%
filter(str_detect(category, "Attendance|Admissions"))
triage <- final_data_long %>% anti_join(attend_admit)
attend_admit_long <-
attend_admit %>%
group_by(row_n) %>%
pivot_wider(id_cols = c(row_n, hospital), names_from = category,
values_from = value)
triage %>%
inner_join(attend_admit_long, by = c("row_n", "hospital")) %>%
arrange(hospital) %>%
select(-row_n)
输出
# A tibble: 90 x 5
hospital category value Attendance Admissions
<chr> <chr> <chr> <chr> <chr>
1 Armadale Tri1 N/A 166 19
2 Armadale Tri2 16 166 19
3 Armadale Tri3 62 166 19
4 Armadale Tri4 79 166 19
5 Armadale Tri5 9 166 19
6 Armadale Tri1 N/A 175 25
7 Armadale Tri2 26 175 25
8 Armadale Tri3 73 175 25
9 Armadale Tri4 55 175 25
10 Armadale Tri5 19 175 25
# … with 80 more rows
数据
*我无法让 OP 的 dput 工作，这里的版本可以是 copy/pasted:
final.data.raw <- structure(
list(RoyalPerth.Attendance = c(235, 209), RoyalPerth.Admissions = c(99, 97), RoyalPerth.Tri1 = c("8", "N/A"),
RoyalPerth.Tri2 = c(33, 41), RoyalPerth.Tri3 = c(89, 73), RoyalPerth.Tri4 = c(85, 80),
RoyalPerth.Tri5 = c("20", "14"), Fremantle.Attendance = c(155, 145), Fremantle.Admissions = c(70, 56),
Fremantle.Tri1 = c("N/A", "N/A"), Fremantle.Tri2 = c(25, 22), Fremantle.Tri3 = c(67, 51),
Fremantle.Tri4 = c(54, 47), Fremantle.Tri5 = c(9, 24), PrincessMargaret.Attendance = c(252, 219),
PrincessMargaret.Admissions = c(59,47), PrincessMargaret.Tri1 = c("N/A", "N/A"), PrincessMargaret.Tri2 = c("13", "14"),
PrincessMargaret.Tri3 = c(75, 61), PrincessMargaret.Tri4 = c(159, 139), PrincessMargaret.Tri5 = c("4", "4"),
KingEdward.Attendance = c(52, 43), KingEdward.Admissions = c("6", "7"), KingEdward.Tri1 = c("N/A", "N/A"),
KingEdward.Tri2 = c("N/A", "N/A"), KingEdward.Tri3 = c("7", "N/A"), KingEdward.Tri4 = c(20, 25),
KingEdward.Tri5 = c("25", "17"), SirCharles.Attendance = c(209, 184), SirCharles.Admissions = c(109, 112),
SirCharles.Tri1 = c("N/A", "N/A"), SirCharles.Tri2 = c(42, 43), SirCharles.Tri3 = c(108, 73),
SirCharles.Tri4 = c(47, 61), SirCharles.Tri5 = c("11", "5"), Armadale.Attendance = c(166, 175),
Armadale.Admissions = c(19, 25), Armadale.Tri1 = c("N/A", "N/A"), Armadale.Tri2 = c(16, 26),
Armadale.Tri3 = c(62, 73), Armadale.Tri4 = c(79, 55), Armadale.Tri5 = c("9", "19"),
Swan.Attendance = c(133, 129), Swan.Admissions = c(17, 25), Swan.Tri1 = c("N/A", "N/A"),
Swan.Tri2 = c(29, 25), Swan.Tri3 = c(59, 57), Swan.Tri4 = c(42, 43),
Swan.Tri5 = c("N/A", "4"), Rockingham.Attendance = c(155, 145), Rockingham.Admissions = c("10", "24"),
Rockingham.Tri1 = c("N/A", "N/A"), Rockingham.Tri2 = c(12, 26), Rockingham.Tri3 = c(51, 45),
Rockingham.Tri4 = c(81, 65), Rockingham.Tri5 = c("11", "8"), Joondalup.Attendance = c(267, 241),
Joondalup.Admissions = c(73, 81), Joondalup.Tri1 = c("N/A", "N/A"), Joondalup.Tri2 = c(27, 23),
Joondalup.Tri3 = c(75, 78), Joondalup.Tri4 = c(151, 133), Joondalup.Tri5 = c("12", "7")),
row.names = 1:2, class = "data.frame")
我正在使用以下数据框。
df2<-final.data%>% gather(Hospital,Attendance,contains("Attendance"))
df2 %>% spread(Hospital, Attendance)
> dput(final.data[0:2,])
structure(list(RoyalPerth.Attendance = c(235, 209), RoyalPerth.Admissions = c(99,
97), RoyalPerth.Tri1 = c("8", "N/A"), RoyalPerth.Tri2 = c(33,
41), RoyalPerth.Tri3 = c(89, 73), RoyalPerth.Tri4 = c(85, 80),
RoyalPert
h.Tri5 = c("20", "14"), Fremantle.Attendance = c(155,
145), Fremantle.Admissions = c(70, 56), Fremantle.Tri1 = c("N/A",
"N/A"), Fremantle.Tri2 = c(25, 22), Fremantle.Tri3 = c(67,
51), Fremantle.Tri4 = c(54, 47), Fremantle.Tri5 = c(9, 24
), PrincessMargaret.Attendance = c(252, 219), PrincessMargaret.Admissions = c(59,
47), PrincessMargaret.Tri1 = c("N/A", "N/A"), PrincessMargaret.Tri2 = c("13",
"14"), PrincessMargaret.Tri3 = c(75, 61), PrincessMargaret.Tri4 = c(159,
139), PrincessMargaret.Tri5 = c("4", "4"), KingEdward.Attendance = c(52,
43), KingEdward.Admissions = c("6", "7"), KingEdward.Tri1 = c("N/A",
"N/A"), KingEdward.Tri2 = c("N/A", "N/A"), KingEdward.Tri3 = c("7",
"N/A"), KingEdward.Tri4 = c(20, 25), KingEdward.Tri5 = c("25",
"17"), SirCharles.Attendance = c(209, 184), SirCharles.Admissions = c(109,
112), SirCharles.Tri1 = c("N/A", "N/A"), SirCharles.Tri2 = c(42,
43), SirCharles.Tri3 = c(108, 73), SirCharles.Tri4 = c(47,
61), SirCharles.Tri5 = c("11", "5"), Armadale.Attendance = c(166,
175), Armadale.Admissions = c(19, 25), Armadale.Tri1 = c("N/A",
"N/A"), Armadale.Tri2 = c(16, 26), Armadale.Tri3 = c(62,
73), Armadale.Tri4 = c(79, 55), Armadale.Tri5 = c("9", "19"
), Swan.Attendance = c(133, 129), Swan.Admissions = c(17,
25), Swan.Tri1 = c("N/A", "N/A"), Swan.Tri2 = c(29, 25),
Swan.Tri3 = c(59, 57), Swan.Tri4 = c(42, 43), Swan.Tri5 = c("N/A",
"4"), Rockingham.Attendance = c(155, 145), Rockingham.Admissions = c("10",
"24"), Rockingham.Tri1 = c("N/A", "N/A"), Rockingham.Tri2 = c(12,
26), Rockingham.Tri3 = c(51, 45), Rockingham.Tri4 = c(81,
65), Rockingham.Tri5 = c("11", "8"), Joondalup.Attendance = c(267,
241), Joondalup.Admissions = c(73, 81), Joondalup.Tri1 = c("N/A",
"N/A"), Joondalup.Tri2 = c(27, 23), Joondalup.Tri3 = c(75,
78), Joondalup.Tri4 = c(151, 133), Joondalup.Tri5 = c("12",
"7")), row.names = 1:2, class = "data.frame")
错误：
警告信息： 测量变量之间的属性不相同； 他们将被删除
我试过以下方法： hospital.dataset<-gather(hospital,triage,sum,Tri1:Tri5) 收集分类 在数据集上使用 cbind 之后。
我想使用 gather 将其转换为长数据集。
dput(hospital.dataset[1:2,])
structure(list(Date = structure(c(-714598, -714597), class = "Date"), [enter image description here][1]
Attendance = c(235, 209), Admissions = c(99, 97), Hospital = structure(c(1L,
1L), .Label = c("RoyalPerth Hospital", "Fremantle Hospital",
"Princess Margaret Hospital", "KingEdward Hospital", "SirCharles Hospital",
"Armadale Hospital", "Swan Hospital", "Rockingham Hospital",
"Joondalup Hospital"), class = "factor"), triage = c("Tri1",
"Tri1"), sum = c(8, 0)), row.names = 1:2, class = "data.frame")
像这样。
提前致谢。
Expected Dataframe
注：这个方案感觉很费力。所以请考虑可能有更优雅的方法可用。
此数据的一个问题是您想要“宽”值（Attendance、Admissions）与您想要“长”值（Tri1、Tri1、 Tri2，等等）。
这个解决方案在整个数据框上使用pivot_longer（注意：pivot_longer是新的gather语法），然后separate拉出医院名称来自特定数据字段。
然后它分成两个数据帧，将pivot_wider应用于Attendance/Admissions列，然后重新加入。
library(tidyverse)
final_data_long <- final.data.raw %>%
mutate_all(as.character) %>%
mutate(row_n = row_number()) %>%
pivot_longer(-row_n, names_to = "field", values_to = "value") %>%
separate(field, into = c("hospital", "category"))
attend_admit <- final_data_long %>%
filter(str_detect(category, "Attendance|Admissions"))
triage <- final_data_long %>% anti_join(attend_admit)
attend_admit_long <-
attend_admit %>%
group_by(row_n) %>%
pivot_wider(id_cols = c(row_n, hospital), names_from = category,
values_from = value)
triage %>%
inner_join(attend_admit_long, by = c("row_n", "hospital")) %>%
arrange(hospital) %>%
select(-row_n)
输出
# A tibble: 90 x 5
hospital category value Attendance Admissions
<chr> <chr> <chr> <chr> <chr>
1 Armadale Tri1 N/A 166 19
2 Armadale Tri2 16 166 19
3 Armadale Tri3 62 166 19
4 Armadale Tri4 79 166 19
5 Armadale Tri5 9 166 19
6 Armadale Tri1 N/A 175 25
7 Armadale Tri2 26 175 25
8 Armadale Tri3 73 175 25
9 Armadale Tri4 55 175 25
10 Armadale Tri5 19 175 25
# … with 80 more rows
数据
*我无法让 OP 的 dput 工作，这里的版本可以是 copy/pasted:
final.data.raw <- structure(
list(RoyalPerth.Attendance = c(235, 209), RoyalPerth.Admissions = c(99, 97), RoyalPerth.Tri1 = c("8", "N/A"),
RoyalPerth.Tri2 = c(33, 41), RoyalPerth.Tri3 = c(89, 73), RoyalPerth.Tri4 = c(85, 80),
RoyalPerth.Tri5 = c("20", "14"), Fremantle.Attendance = c(155, 145), Fremantle.Admissions = c(70, 56),
Fremantle.Tri1 = c("N/A", "N/A"), Fremantle.Tri2 = c(25, 22), Fremantle.Tri3 = c(67, 51),
Fremantle.Tri4 = c(54, 47), Fremantle.Tri5 = c(9, 24), PrincessMargaret.Attendance = c(252, 219),
PrincessMargaret.Admissions = c(59,47), PrincessMargaret.Tri1 = c("N/A", "N/A"), PrincessMargaret.Tri2 = c("13", "14"),
PrincessMargaret.Tri3 = c(75, 61), PrincessMargaret.Tri4 = c(159, 139), PrincessMargaret.Tri5 = c("4", "4"),
KingEdward.Attendance = c(52, 43), KingEdward.Admissions = c("6", "7"), KingEdward.Tri1 = c("N/A", "N/A"),
KingEdward.Tri2 = c("N/A", "N/A"), KingEdward.Tri3 = c("7", "N/A"), KingEdward.Tri4 = c(20, 25),
KingEdward.Tri5 = c("25", "17"), SirCharles.Attendance = c(209, 184), SirCharles.Admissions = c(109, 112),
SirCharles.Tri1 = c("N/A", "N/A"), SirCharles.Tri2 = c(42, 43), SirCharles.Tri3 = c(108, 73),
SirCharles.Tri4 = c(47, 61), SirCharles.Tri5 = c("11", "5"), Armadale.Attendance = c(166, 175),
Armadale.Admissions = c(19, 25), Armadale.Tri1 = c("N/A", "N/A"), Armadale.Tri2 = c(16, 26),
Armadale.Tri3 = c(62, 73), Armadale.Tri4 = c(79, 55), Armadale.Tri5 = c("9", "19"),
Swan.Attendance = c(133, 129), Swan.Admissions = c(17, 25), Swan.Tri1 = c("N/A", "N/A"),
Swan.Tri2 = c(29, 25), Swan.Tri3 = c(59, 57), Swan.Tri4 = c(42, 43),
Swan.Tri5 = c("N/A", "4"), Rockingham.Attendance = c(155, 145), Rockingham.Admissions = c("10", "24"),
Rockingham.Tri1 = c("N/A", "N/A"), Rockingham.Tri2 = c(12, 26), Rockingham.Tri3 = c(51, 45),
Rockingham.Tri4 = c(81, 65), Rockingham.Tri5 = c("11", "8"), Joondalup.Attendance = c(267, 241),
Joondalup.Admissions = c(73, 81), Joondalup.Tri1 = c("N/A", "N/A"), Joondalup.Tri2 = c(27, 23),
Joondalup.Tri3 = c(75, 78), Joondalup.Tri4 = c(151, 133), Joondalup.Tri5 = c("12", "7")),
row.names = 1:2, class = "data.frame")