时间戳

首页 > 解决方案 > 有条件地对数据帧的每一行进行计数

r - 有条件地对数据帧的每一行进行计数

问题描述

我有一个包含 97 个变量的数据框。对于变量 32 到 97，我需要计算每行中出现的值 1、2、3 和 4 的实例数（仅适用于从 32 到 97 的变量）。

下面是数据框的结构：

 structure(list(subject_label = c("01", "01"), 
 event_code = c("b", "16"), site_id = c(5, 5), site_label = c("a", 
 "a"), event_label = c("Baseline", 
 NA), done = c(1L, NA), ndreason = c(NA_integer_, NA_integer_
 ), rater = c("OLD", NA), carerid = c(1L, NA), ntgph = c(4L, 
 NA), nt1ph = c(2L, NA), ntgmh = c(2L, NA), ntgse = c(0L, 
 NA), ntgxv = c(1L, NA), ntgxb = c(0L, NA), ntgxg = c(0L, 
 NA), ntgxh = c(0L, NA), ntgxd = c(0L, NA), ntgxa = c(0L, 
 NA), ntgxm = c(0L, NA), ntgxw = c(0L, NA), ntgxp = c(0L, 
 NA), ntgcd = c(0L, NA), ntgch = c(0L, NA), ntgcs = c(0L, 
 NA), ntgnr = c(0L, NA), ntgii = c(0L, NA), ntsem = c(0L, 
 NA), ntsei = c(0L, NA), ntsev = c(0L, NA), ntsey = c(NA_integer_, 
 NA_integer_), ntseo = c("Interpersonal conflict with co-workers", 
 NA), ntawb = c(0L, NA), ntahd = c(0L, NA), ntaid = c(0L, 
 NA), ntaui = c(0L, NA), ntahe = c(0L, NA), ntaub = c(0L, 
 NA), ntadi = c(1L, NA), ntatb = c(NA_integer_, NA_integer_
 ), ntcic = c(0L, NA), ntcfw = c(0L, NA), ntcfi = c(0L, NA
 ), ntclc = c(0L, NA), ntcnr = c(0L, NA), ntcnw = c(0L, NA
 ), ntses = c(0L, NA), ntsia = c(0L, NA), ntswf = c(1L, NA
 ), ntscn = c(0L, NA), ntssd = c(0L, NA), ntswn = c(0L, NA
 ), ntswe = c(0L, NA), ntssl = c(0L, NA), ntacw = c(0L, NA
 ), ntalb = c(0L, NA), ntacs = c(NA_integer_, NA_integer_), 
 ntafa = c(0L, NA), ntaaw = c(0L, NA), ntmrp = c(0L, NA), 
 ntmdr = c(0L, NA), ntmre = c(0L, NA), ntmfw = c(0L, NA), 
 ntmlt = c(4L, NA), ntmmo = c(0L, NA), ntmwp = c(0L, NA), 
 ntmpp = c(0L, NA), ntmpl = c(0L, NA), ntbaw = c(0L, NA), 
 ntbws = c(1L, NA), ntbwp = c(4L, NA), ntbla = c(0L, NA), 
 ntbow = c(0L, NA), ntbor = c(4L, NA), ntbhh = c(4L, NA), 
 ntbfo = c(0L, NA), ntbii = c(4L, NA), ntblc = c(0L, NA), 
 ntbaa = c(0L, NA), ntbad = c(0L, NA), ntbva = c(0L, NA), 
 ntbpa = c(0L, NA), ntbtt = c(0L, NA), ntbll = c(0L, NA), 
 ntbts = c(1L, NA), ntrca = c(0L, NA), ntrht = c(0L, NA), 
 ntrst = c(0L, NA), ntrct = c(0L, NA), ntrci = c(1L, NA), 
 ntrcm = c(0L, NA), ntoog = c(0L, NA), ntoop = c(0L, NA), 
 ntoof = c(0L, NA), ntooa = c(0L, NA), ntoow = c(0L, NA), 
 ntoov = c(0L, NA), ntgsp = c("a", 
 NA), update_stamp = c("a", NA), class = "data.frame")

我想做的是：

 NTG$symptoms_morethanyear  <- length(which(NTG[,c(32:97)] == 1))

然而，这似乎总结了这些字段中的所有内容并将整个总和放在最后一列中（不是每行单独进行）。

标签： rdplyrcountsumconditional-statements

解决方案

如果您不介意将第 37 列修改为第 97 列，这是一个解决方案

library(tidyverse)


data_frame%>%
   mutate_at(vars(37:97), list(~if_else(. %in% 1:4, 1, 0)))%>%
   mutate(conts = rowSums(select(., 37:97)))%>%select(97:101)

  ntoov ntgsp update_stamp   conts
1     0     a            a     10
2     0  <NA>         <NA>      0

如果您不想修改第 32 到 97 列，也可以这样做：

library(data.table)

data_frame$counts <- rowSums(data_frame[,32:97] %between% c(1,4), na.rm=T)

data_frame%>%select(90:101)

  ntrci ntrcm ntoog ntoop ntoof ntooa ntoow ntoov ntgsp update_stamp   counts
1     1     0     0     0     0     0     0     0     a            a     10
2    NA    NA    NA    NA    NA    NA    NA    NA  <NA>         <NA>     0

推荐阅读