转置数据帧
Transposing data frames
周末快乐。
我一直在尝试在 R 中复制此 blog post 的结果。我正在寻找一种不使用 t 转置数据的方法，最好使用 tidyr 或reshape。在下面的示例中，metadata 是通过转置 data.
获得的
metadata <- data.frame(colnames(data), t(data[1:4, ]) )
colnames(metadata) <- t(metadata[1,])
metadata <- metadata[-1,]
metadata$Multiplier <- as.numeric(metadata$Multiplier)
虽然达到了我的要求，但我觉得有点不熟练。是否有任何有效的工作流程来转置数据框？
数据输入
data <- structure(list(Series.Description = c("Unit:", "Multiplier:",
"Currency:", "Unique Identifier: "), Nominal.Broad.Dollar.Index. = c("Index:_1997_Jan_100",
"1", NA, "H10/H10/JRXWTFB_N.M"), Nominal.Major.Currencies.Dollar.Index. = c("Index:_1973_Mar_100",
"1", NA, "H10/H10/JRXWTFN_N.M"), Nominal.Other.Important.Trading.Partners.Dollar.Index. = c("Index:_1997_Jan_100",
"1", NA, "H10/H10/JRXWTFO_N.M"), AUSTRALIA....SPOT.EXCHANGE.RATE..US..AUSTRALIAN...RECIPROCAL.OF.RXI_N.M.AL. = c("Currency:_Per_AUD",
"1", "USD", "H10/H10/RXI$US_N.M.AL"), SPOT.EXCHANGE.RATE...EURO.AREA. = c("Currency:_Per_EUR",
"1", "USD", "H10/H10/RXI$US_N.M.EU"), NEW.ZEALAND....SPOT.EXCHANGE.RATE..US..NZ...RECIPROCAL.OF.RXI_N.M.NZ.. = c("Currency:_Per_NZD",
"1", "USD", "H10/H10/RXI$US_N.M.NZ"), United.Kingdom....Spot.Exchange.Rate..US..Pound.Sterling.Reciprocal.of.rxi_n.m.uk = c("Currency:_Per_GBP",
"0.01", "USD", "H10/H10/RXI$US_N.M.UK"), BRAZIL....SPOT.EXCHANGE.RATE..REAIS.US.. = c("Currency:_Per_USD",
"1", "BRL", "H10/H10/RXI_N.M.BZ"), CANADA....SPOT.EXCHANGE.RATE..CANADIAN...US.. = c("Currency:_Per_USD",
"1", "CAD", "H10/H10/RXI_N.M.CA"), CHINA....SPOT.EXCHANGE.RATE..YUAN.US.. = c("Currency:_Per_USD",
"1", "CNY", "H10/H10/RXI_N.M.CH"), DENMARK....SPOT.EXCHANGE.RATE..KRONER.US.. = c("Currency:_Per_USD",
"1", "DKK", "H10/H10/RXI_N.M.DN"), HONG.KONG....SPOT.EXCHANGE.RATE..HK..US.. = c("Currency:_Per_USD",
"1", "HKD", "H10/H10/RXI_N.M.HK"), INDIA....SPOT.EXCHANGE.RATE..RUPEES.US. = c("Currency:_Per_USD",
"1", "INR", "H10/H10/RXI_N.M.IN"), JAPAN....SPOT.EXCHANGE.RATE..YEA.US.. = c("Currency:_Per_USD",
"1", "JPY", "H10/H10/RXI_N.M.JA"), KOREA....SPOT.EXCHANGE.RATE..WON.US.. = c("Currency:_Per_USD",
"1", "KRW", "H10/H10/RXI_N.M.KO"), Malaysia...Spot.Exchange.Rate..Ringgit.US.. = c("Currency:_Per_USD",
"1", "MYR", "H10/H10/RXI_N.M.MA"), MEXICO....SPOT.EXCHANGE.RATE..PESOS.US.. = c("Currency:_Per_USD",
"1", "MXN", "H10/H10/RXI_N.M.MX"), NORWAY....SPOT.EXCHANGE.RATE..KRONER.US.. = c("Currency:_Per_USD",
"1", "NOK", "H10/H10/RXI_N.M.NO"), SWEDEN....SPOT.EXCHANGE.RATE..KRONOR.US.. = c("Currency:_Per_USD",
"1", "SEK", "H10/H10/RXI_N.M.SD"), SOUTH.AFRICA....SPOT.EXCHANGE.RATE..RAND.US.. = c("Currency:_Per_USD",
"1", "ZAR", "H10/H10/RXI_N.M.SF"), Singapore...SPOT.EXCHANGE.RATE..SINGAPORE...US.. = c("Currency:_Per_USD",
"1", "SGD", "H10/H10/RXI_N.M.SI"), SRI.LANKA....SPOT.EXCHANGE.RATE..RUPEES.US.. = c("Currency:_Per_USD",
"1", "LKR", "H10/H10/RXI_N.M.SL"), SWITZERLAND....SPOT.EXCHANGE.RATE..FRANCS.US.. = c("Currency:_Per_USD",
"1", "CHF", "H10/H10/RXI_N.M.SZ"), TAIWAN....SPOT.EXCHANGE.RATE..NT..US.. = c("Currency:_Per_USD",
"1", "TWD", "H10/H10/RXI_N.M.TA"), THAILAND....SPOT.EXCHANGE.RATE....THAILAND. = c("Currency:_Per_USD",
"1", "THB", "H10/H10/RXI_N.M.TH"), VENEZUELA....SPOT.EXCHANGE.RATE..BOLIVARES.US.. = c("Currency:_Per_USD",
"1", "VEB", "H10/H10/RXI_N.M.VE")), .Names = c("Series.Description",
"Nominal.Broad.Dollar.Index.", "Nominal.Major.Currencies.Dollar.Index.",
"Nominal.Other.Important.Trading.Partners.Dollar.Index.", "AUSTRALIA....SPOT.EXCHANGE.RATE..US..AUSTRALIAN...RECIPROCAL.OF.RXI_N.M.AL.",
"SPOT.EXCHANGE.RATE...EURO.AREA.", "NEW.ZEALAND....SPOT.EXCHANGE.RATE..US..NZ...RECIPROCAL.OF.RXI_N.M.NZ..",
"United.Kingdom....Spot.Exchange.Rate..US..Pound.Sterling.Reciprocal.of.rxi_n.m.uk",
"BRAZIL....SPOT.EXCHANGE.RATE..REAIS.US..", "CANADA....SPOT.EXCHANGE.RATE..CANADIAN...US..",
"CHINA....SPOT.EXCHANGE.RATE..YUAN.US..", "DENMARK....SPOT.EXCHANGE.RATE..KRONER.US..",
"HONG.KONG....SPOT.EXCHANGE.RATE..HK..US..", "INDIA....SPOT.EXCHANGE.RATE..RUPEES.US.",
"JAPAN....SPOT.EXCHANGE.RATE..YEA.US..", "KOREA....SPOT.EXCHANGE.RATE..WON.US..",
"Malaysia...Spot.Exchange.Rate..Ringgit.US..", "MEXICO....SPOT.EXCHANGE.RATE..PESOS.US..",
"NORWAY....SPOT.EXCHANGE.RATE..KRONER.US..", "SWEDEN....SPOT.EXCHANGE.RATE..KRONOR.US..",
"SOUTH.AFRICA....SPOT.EXCHANGE.RATE..RAND.US..", "Singapore...SPOT.EXCHANGE.RATE..SINGAPORE...US..",
"SRI.LANKA....SPOT.EXCHANGE.RATE..RUPEES.US..", "SWITZERLAND....SPOT.EXCHANGE.RATE..FRANCS.US..",
"TAIWAN....SPOT.EXCHANGE.RATE..NT..US..", "THAILAND....SPOT.EXCHANGE.RATE....THAILAND.",
"VENEZUELA....SPOT.EXCHANGE.RATE..BOLIVARES.US.."), row.names = c(NA,
4L), class = "data.frame")
使用 tidyr，你 gather 除了第一列之外的所有列，然后你 spread 收集的列。
尝试：
library(dplyr)
library(tidyr)
data %>%
gather(var, val, 2:ncol(data)) %>%
spread(Series.Description, val)
library(dplyr)
# Omitted data <- structure part ...
这是复制主要答案中的内容的内容，但更通用（例如，在 Series.Description 不是结果的第一列的情况下工作）并使用较新的 pivot_wider/pivot_longer 动词.
df_transpose <- function(df) {
df %>%
tidyr::pivot_longer(-1) %>%
tidyr::pivot_wider(names_from = 1, values_from = value)
}
df_transpose(data)
#> # A tibble: 26 x 5
#> name `Unit:` `Multiplier:` `Currency:` `Unique Identifi…
#> <chr> <chr> <chr> <chr> <chr>
#> 1 Nominal.Broad.Dollar.… Index:_19… 1 <NA> H10/H10/JRXWTFB_…
#> 2 Nominal.Major.Currenc… Index:_19… 1 <NA> H10/H10/JRXWTFN_…
#> 3 Nominal.Other.Importa… Index:_19… 1 <NA> H10/H10/JRXWTFO_…
#> 4 AUSTRALIA....SPOT.EXC… Currency:… 1 USD H10/H10/RXI$US_N…
#> 5 SPOT.EXCHANGE.RATE...… Currency:… 1 USD H10/H10/RXI$US_N…
#> 6 NEW.ZEALAND....SPOT.E… Currency:… 1 USD H10/H10/RXI$US_N…
#> 7 United.Kingdom....Spo… Currency:… 0.01 USD H10/H10/RXI$US_N…
#> 8 BRAZIL....SPOT.EXCHAN… Currency:… 1 BRL H10/H10/RXI_N.M.…
#> 9 CANADA....SPOT.EXCHAN… Currency:… 1 CAD H10/H10/RXI_N.M.…
#> 10 CHINA....SPOT.EXCHANG… Currency:… 1 CNY H10/H10/RXI_N.M.…
#> # … with 16 more rows
但请注意（如上面的答案）第一列的名称丢失了。以下内容保留了这一点（我想上面@jbkunst 提出的 spread_(names(data)[1], "val") 方法也是如此）。
df_transpose <- function(df) {
first_name <- colnames(df)[1]
temp <-
df %>%
tidyr::pivot_longer(-1) %>%
tidyr::pivot_wider(names_from = 1, values_from = value)
colnames(temp)[1] <- first_name
temp
}
df_transpose(data)
#> # A tibble: 26 x 5
#> Series.Description `Unit:` `Multiplier:` `Currency:` `Unique Identif…
#> <chr> <chr> <chr> <chr> <chr>
#> 1 Nominal.Broad.Dollar.In… Index:_1… 1 <NA> H10/H10/JRXWTFB…
#> 2 Nominal.Major.Currencie… Index:_1… 1 <NA> H10/H10/JRXWTFN…
#> 3 Nominal.Other.Important… Index:_1… 1 <NA> H10/H10/JRXWTFO…
#> 4 AUSTRALIA....SPOT.EXCHA… Currency… 1 USD H10/H10/RXI$US_…
#> 5 SPOT.EXCHANGE.RATE...EU… Currency… 1 USD H10/H10/RXI$US_…
#> 6 NEW.ZEALAND....SPOT.EXC… Currency… 1 USD H10/H10/RXI$US_…
#> 7 United.Kingdom....Spot.… Currency… 0.01 USD H10/H10/RXI$US_…
#> 8 BRAZIL....SPOT.EXCHANGE… Currency… 1 BRL H10/H10/RXI_N.M…
#> 9 CANADA....SPOT.EXCHANGE… Currency… 1 CAD H10/H10/RXI_N.M…
#> 10 CHINA....SPOT.EXCHANGE.… Currency… 1 CNY H10/H10/RXI_N.M…
#> # … with 16 more rows
由 reprex package (v2.0.0)
于 2021-05-30 创建
周末快乐。
我一直在尝试在 R 中复制此 blog post 的结果。我正在寻找一种不使用 t 转置数据的方法，最好使用 tidyr 或reshape。在下面的示例中，metadata 是通过转置 data.
metadata <- data.frame(colnames(data), t(data[1:4, ]) )
colnames(metadata) <- t(metadata[1,])
metadata <- metadata[-1,]
metadata$Multiplier <- as.numeric(metadata$Multiplier)
虽然达到了我的要求，但我觉得有点不熟练。是否有任何有效的工作流程来转置数据框？
数据输入
data <- structure(list(Series.Description = c("Unit:", "Multiplier:",
"Currency:", "Unique Identifier: "), Nominal.Broad.Dollar.Index. = c("Index:_1997_Jan_100",
"1", NA, "H10/H10/JRXWTFB_N.M"), Nominal.Major.Currencies.Dollar.Index. = c("Index:_1973_Mar_100",
"1", NA, "H10/H10/JRXWTFN_N.M"), Nominal.Other.Important.Trading.Partners.Dollar.Index. = c("Index:_1997_Jan_100",
"1", NA, "H10/H10/JRXWTFO_N.M"), AUSTRALIA....SPOT.EXCHANGE.RATE..US..AUSTRALIAN...RECIPROCAL.OF.RXI_N.M.AL. = c("Currency:_Per_AUD",
"1", "USD", "H10/H10/RXI$US_N.M.AL"), SPOT.EXCHANGE.RATE...EURO.AREA. = c("Currency:_Per_EUR",
"1", "USD", "H10/H10/RXI$US_N.M.EU"), NEW.ZEALAND....SPOT.EXCHANGE.RATE..US..NZ...RECIPROCAL.OF.RXI_N.M.NZ.. = c("Currency:_Per_NZD",
"1", "USD", "H10/H10/RXI$US_N.M.NZ"), United.Kingdom....Spot.Exchange.Rate..US..Pound.Sterling.Reciprocal.of.rxi_n.m.uk = c("Currency:_Per_GBP",
"0.01", "USD", "H10/H10/RXI$US_N.M.UK"), BRAZIL....SPOT.EXCHANGE.RATE..REAIS.US.. = c("Currency:_Per_USD",
"1", "BRL", "H10/H10/RXI_N.M.BZ"), CANADA....SPOT.EXCHANGE.RATE..CANADIAN...US.. = c("Currency:_Per_USD",
"1", "CAD", "H10/H10/RXI_N.M.CA"), CHINA....SPOT.EXCHANGE.RATE..YUAN.US.. = c("Currency:_Per_USD",
"1", "CNY", "H10/H10/RXI_N.M.CH"), DENMARK....SPOT.EXCHANGE.RATE..KRONER.US.. = c("Currency:_Per_USD",
"1", "DKK", "H10/H10/RXI_N.M.DN"), HONG.KONG....SPOT.EXCHANGE.RATE..HK..US.. = c("Currency:_Per_USD",
"1", "HKD", "H10/H10/RXI_N.M.HK"), INDIA....SPOT.EXCHANGE.RATE..RUPEES.US. = c("Currency:_Per_USD",
"1", "INR", "H10/H10/RXI_N.M.IN"), JAPAN....SPOT.EXCHANGE.RATE..YEA.US.. = c("Currency:_Per_USD",
"1", "JPY", "H10/H10/RXI_N.M.JA"), KOREA....SPOT.EXCHANGE.RATE..WON.US.. = c("Currency:_Per_USD",
"1", "KRW", "H10/H10/RXI_N.M.KO"), Malaysia...Spot.Exchange.Rate..Ringgit.US.. = c("Currency:_Per_USD",
"1", "MYR", "H10/H10/RXI_N.M.MA"), MEXICO....SPOT.EXCHANGE.RATE..PESOS.US.. = c("Currency:_Per_USD",
"1", "MXN", "H10/H10/RXI_N.M.MX"), NORWAY....SPOT.EXCHANGE.RATE..KRONER.US.. = c("Currency:_Per_USD",
"1", "NOK", "H10/H10/RXI_N.M.NO"), SWEDEN....SPOT.EXCHANGE.RATE..KRONOR.US.. = c("Currency:_Per_USD",
"1", "SEK", "H10/H10/RXI_N.M.SD"), SOUTH.AFRICA....SPOT.EXCHANGE.RATE..RAND.US.. = c("Currency:_Per_USD",
"1", "ZAR", "H10/H10/RXI_N.M.SF"), Singapore...SPOT.EXCHANGE.RATE..SINGAPORE...US.. = c("Currency:_Per_USD",
"1", "SGD", "H10/H10/RXI_N.M.SI"), SRI.LANKA....SPOT.EXCHANGE.RATE..RUPEES.US.. = c("Currency:_Per_USD",
"1", "LKR", "H10/H10/RXI_N.M.SL"), SWITZERLAND....SPOT.EXCHANGE.RATE..FRANCS.US.. = c("Currency:_Per_USD",
"1", "CHF", "H10/H10/RXI_N.M.SZ"), TAIWAN....SPOT.EXCHANGE.RATE..NT..US.. = c("Currency:_Per_USD",
"1", "TWD", "H10/H10/RXI_N.M.TA"), THAILAND....SPOT.EXCHANGE.RATE....THAILAND. = c("Currency:_Per_USD",
"1", "THB", "H10/H10/RXI_N.M.TH"), VENEZUELA....SPOT.EXCHANGE.RATE..BOLIVARES.US.. = c("Currency:_Per_USD",
"1", "VEB", "H10/H10/RXI_N.M.VE")), .Names = c("Series.Description",
"Nominal.Broad.Dollar.Index.", "Nominal.Major.Currencies.Dollar.Index.",
"Nominal.Other.Important.Trading.Partners.Dollar.Index.", "AUSTRALIA....SPOT.EXCHANGE.RATE..US..AUSTRALIAN...RECIPROCAL.OF.RXI_N.M.AL.",
"SPOT.EXCHANGE.RATE...EURO.AREA.", "NEW.ZEALAND....SPOT.EXCHANGE.RATE..US..NZ...RECIPROCAL.OF.RXI_N.M.NZ..",
"United.Kingdom....Spot.Exchange.Rate..US..Pound.Sterling.Reciprocal.of.rxi_n.m.uk",
"BRAZIL....SPOT.EXCHANGE.RATE..REAIS.US..", "CANADA....SPOT.EXCHANGE.RATE..CANADIAN...US..",
"CHINA....SPOT.EXCHANGE.RATE..YUAN.US..", "DENMARK....SPOT.EXCHANGE.RATE..KRONER.US..",
"HONG.KONG....SPOT.EXCHANGE.RATE..HK..US..", "INDIA....SPOT.EXCHANGE.RATE..RUPEES.US.",
"JAPAN....SPOT.EXCHANGE.RATE..YEA.US..", "KOREA....SPOT.EXCHANGE.RATE..WON.US..",
"Malaysia...Spot.Exchange.Rate..Ringgit.US..", "MEXICO....SPOT.EXCHANGE.RATE..PESOS.US..",
"NORWAY....SPOT.EXCHANGE.RATE..KRONER.US..", "SWEDEN....SPOT.EXCHANGE.RATE..KRONOR.US..",
"SOUTH.AFRICA....SPOT.EXCHANGE.RATE..RAND.US..", "Singapore...SPOT.EXCHANGE.RATE..SINGAPORE...US..",
"SRI.LANKA....SPOT.EXCHANGE.RATE..RUPEES.US..", "SWITZERLAND....SPOT.EXCHANGE.RATE..FRANCS.US..",
"TAIWAN....SPOT.EXCHANGE.RATE..NT..US..", "THAILAND....SPOT.EXCHANGE.RATE....THAILAND.",
"VENEZUELA....SPOT.EXCHANGE.RATE..BOLIVARES.US.."), row.names = c(NA,
4L), class = "data.frame")
使用 tidyr，你 gather 除了第一列之外的所有列，然后你 spread 收集的列。
尝试：
library(dplyr)
library(tidyr)
data %>%
gather(var, val, 2:ncol(data)) %>%
spread(Series.Description, val)
library(dplyr)
# Omitted data <- structure part ...
这是复制主要答案中的内容的内容，但更通用（例如，在 Series.Description 不是结果的第一列的情况下工作）并使用较新的 pivot_wider/pivot_longer 动词.
df_transpose <- function(df) {
df %>%
tidyr::pivot_longer(-1) %>%
tidyr::pivot_wider(names_from = 1, values_from = value)
}
df_transpose(data)
#> # A tibble: 26 x 5
#> name `Unit:` `Multiplier:` `Currency:` `Unique Identifi…
#> <chr> <chr> <chr> <chr> <chr>
#> 1 Nominal.Broad.Dollar.… Index:_19… 1 <NA> H10/H10/JRXWTFB_…
#> 2 Nominal.Major.Currenc… Index:_19… 1 <NA> H10/H10/JRXWTFN_…
#> 3 Nominal.Other.Importa… Index:_19… 1 <NA> H10/H10/JRXWTFO_…
#> 4 AUSTRALIA....SPOT.EXC… Currency:… 1 USD H10/H10/RXI$US_N…
#> 5 SPOT.EXCHANGE.RATE...… Currency:… 1 USD H10/H10/RXI$US_N…
#> 6 NEW.ZEALAND....SPOT.E… Currency:… 1 USD H10/H10/RXI$US_N…
#> 7 United.Kingdom....Spo… Currency:… 0.01 USD H10/H10/RXI$US_N…
#> 8 BRAZIL....SPOT.EXCHAN… Currency:… 1 BRL H10/H10/RXI_N.M.…
#> 9 CANADA....SPOT.EXCHAN… Currency:… 1 CAD H10/H10/RXI_N.M.…
#> 10 CHINA....SPOT.EXCHANG… Currency:… 1 CNY H10/H10/RXI_N.M.…
#> # … with 16 more rows
但请注意（如上面的答案）第一列的名称丢失了。以下内容保留了这一点（我想上面@jbkunst 提出的 spread_(names(data)[1], "val") 方法也是如此）。
df_transpose <- function(df) {
first_name <- colnames(df)[1]
temp <-
df %>%
tidyr::pivot_longer(-1) %>%
tidyr::pivot_wider(names_from = 1, values_from = value)
colnames(temp)[1] <- first_name
temp
}
df_transpose(data)
#> # A tibble: 26 x 5
#> Series.Description `Unit:` `Multiplier:` `Currency:` `Unique Identif…
#> <chr> <chr> <chr> <chr> <chr>
#> 1 Nominal.Broad.Dollar.In… Index:_1… 1 <NA> H10/H10/JRXWTFB…
#> 2 Nominal.Major.Currencie… Index:_1… 1 <NA> H10/H10/JRXWTFN…
#> 3 Nominal.Other.Important… Index:_1… 1 <NA> H10/H10/JRXWTFO…
#> 4 AUSTRALIA....SPOT.EXCHA… Currency… 1 USD H10/H10/RXI$US_…
#> 5 SPOT.EXCHANGE.RATE...EU… Currency… 1 USD H10/H10/RXI$US_…
#> 6 NEW.ZEALAND....SPOT.EXC… Currency… 1 USD H10/H10/RXI$US_…
#> 7 United.Kingdom....Spot.… Currency… 0.01 USD H10/H10/RXI$US_…
#> 8 BRAZIL....SPOT.EXCHANGE… Currency… 1 BRL H10/H10/RXI_N.M…
#> 9 CANADA....SPOT.EXCHANGE… Currency… 1 CAD H10/H10/RXI_N.M…
#> 10 CHINA....SPOT.EXCHANGE.… Currency… 1 CNY H10/H10/RXI_N.M…
#> # … with 16 more rows
由 reprex package (v2.0.0)
于 2021-05-30 创建