※ 引述《celestialgod (天)》之铭言： : ※ 引述《lovedmagic (EricZou)》之铭言： : : [问题类型]: : : 程式谘询(我想用R 做某件事情，但是我不知道要怎麽用R 写出来) : : [软体熟悉度]: : : 新手(没写过程式，R 是我的第一次) : : [问题叙述]: : : 目前我遇到的困难是我有A1~A270个变数，我每个变数都有大概五个数值， : : 每个数值左侧代表一个名称，例如我A1这个变数里面有 34 18 16 9 4五个值 : : 然後每个数值对应到左侧的名称有A B C D E ，我想做的就是取出A1变数中 : : 最大的那个值，然後再取出左侧的ID，主要是要求出左侧的ID，我要如何 : : 让回圈把每个变数都跑过一次呢??? : : 以下有错误范例 : : [程式范例]: : : fxck = data.frame() : : for (i in 1:270){ : : number[data$Ai==max(data$Ai)] : : fxck <- rbind(number[Ai==max(Ai)],Ai) : : print(fxck) : : } : : [关键字]: : : 多个变数、回圈 : library(data.table) : DT <- data.table(name = LETTERS[1L:5L], matrix(rnorm(270L * 5L), 5L)) : setnames(DT, paste0("V", 1L:270L), paste0("A", 1L:270L)) : # direct: : data.table( : variable = paste0("A", 1L:270L), : ID = as.character(DT[ , lapply(.SD, function(x) name[which.max(x)]), : .SDcols = A1:A270]) : value = as.numeric(DT[ , lapply(.SD, max), .SDcols = A1:A270]) : ) : # variable ID value : # 1: A1 B 0.5974676 : # 2: A2 A 1.6205859 : # 3: A3 E 0.9247647 : # 4: A4 A 1.7180776 : # 5: A5 D 1.2060909 : # --- : # 266: A266 C 0.4813518 : # 267: A267 B 1.3096330 : # 268: A268 B 0.4504671 : # 269: A269 C 0.7401682 : # 270: A270 D 2.9256539 : # melt: : library(pipeR) : out <- melt.data.table(DT, 1) %>>% : `[`( , .(ID = name[which.max(value)], value = max(value)) , : by =.(variable)) : # variable ID value : # 1: A1 B 0.5974676 : # 2: A2 A 1.6205859 : # 3: A3 E 0.9247647 : # 4: A4 A 1.7180776 : # 5: A5 D 1.2060909 : # --- : # 266: A266 C 0.4813518 : # 267: A267 B 1.3096330 : # 268: A268 B 0.4504671 : # 269: A269 C 0.7401682 : # 270: A270 D 2.9256539 : # melt 2: : out <- melt.data.table(DT, 1) %>>% : `[`( , `:=`(maxRow = rank(value) == length(value)), by = .(variable)) %>>% : `[`(maxRow == TRUE) %>>% `[`( , maxRow := NULL) : # tidyr + dplyr : library(tidyr) : library(dplyr) : library(pipeR) : gather(DT, variable, value, -name) %>>% group_by(variable) %>>% : filter(rank(value) == length(value)) 可能是因为我表达得不够清楚导致讨论的方向不够明确我在这里说声抱歉， 以下是我头脑清醒後想出来的data的处理方式 X A1 A2 A3 1 member A 34 10 4 2 member B 18 90 64 3 member C 16 155 0 4 member D 9 202 1 5 member E 4 84 6 6 member F 0 50 4 7 member G 0 153 30 8 member H 0 0 30 9 member I 0 0 40 10 member J 0 0 14 11 member K 0 0 91 12 member L 0 0 0 13 member M 0 0 0 14 member N 0 0 0 15 member O 0 0 0 16 member P 0 0 0 17 member Q 0 0 0 18 member R 0 0 0 19 member S 0 0 0 20 member T 0 0 0 21 member U 0 0 0 22 member V 0 0 0 23 member W 0 0 0 24 member X 0 0 0 25 member Y 0 0 0 26 member Z 0 0 0 27 member AA 0 0 0 我主要是想求在我变数下最大的数对应到左边的名称 例如A1是member A 然後值是34接着A2是member D然後值是202 以下是我真正想要的解，个人写的步骤有点诡异还会跑出warnings 但总归是跑了理想的长相出来所以跟大家分享一下 A <- cbind() B <- cbind() for (i in 2:271){ colnames <- as.vector(X[data[,i] == max(data[,i])]) A[i-1] <- colnames B[i-1] <- max(data[,i]) } AB <- cbind(A,B) AB A B [1,] "member A" "34" [2,] "member D" "202" [3,] "member K" "91" [4,] "member J" "56" [5,] "member E" "298" [6,] "member B" "318" 以上谢谢之前各位大大们的回答 --

※ 发信站: 批踢踢实业坊(ptt.cc), 来自: 114.47.59.17 ※ 文章网址: https://webptt.com/cn.aspx?n=bbs/R_Language/M.1482725997.A.669.html