在R中按每组变异 [英] mutate per group by in R
问题描述
我想从传感器数据中识别出零件,并给它们一个ID.因此,我想按传感器"列对以下数据集进行分组,并查看值"行是否从0切换为1.确定第一行时,caseid切换为1(如手工列caseid一样).只要值保持为1,它就保持为1.当它变为0时,应切换回0.在下一次从0切换到1的情况下,caseid应该变为2,因为第二部分被传感器识别,依此类推.
i want to identify pieces from sensor data and give them an ID. Therefore I want to group the following dataset by Sensor column and look whether the Value row switched from 0 to 1. When it does the first piece is identified and the caseid switches to 1 (as in the handmade column caseid). It remains 1 as long the value stays 1. When it becomes 0 it should switche back to 0. At the next switch from 0 to 1 the caseid should become 2 because the second piece is recognized by the sensor and so forth..
time = c("07:00:01","07:00:01","07:00:01","07:00:02","07:00:02","07:00:02","07:00:03","07:00:03","07:00:03","07:00:04",
"07:00:04","07:00:04","07:00:05","07:00:05","07:00:05","07:00:06","07:00:06","07:00:06","07:00:07","07:00:07",
"07:00:07","07:00:08","07:00:08","07:00:08","07:00:09","07:00:09","07:00:09")
sensor = c(10001,10002,10003,10001,10002,10003,10001,10002,10003,10001,10002,10003,10001,10002,10003,10001,10002,10003,
10001,10002,10003,10001,10002,10003,10001,10002,10003)
values = c(0,0,0,1,0,0,1,0,0,0,1,0,0,1,0,0,1,0,0,0,0,1,0,1,1,0,1)
caseid = c(0,0,0,1,0,0,1,0,0,0,1,0,0,1,0,0,1,0,0,0,0,2,0,1,2,0,1)
data = data.frame(time,sensor,values,caseid)
(所以data $ caseid是我想要得到的)
(So the data$caseid is what I am trying to get)
我认为可以通过某种方式来实现这一目标,但是我做得不好,因此我选择了另一种(草率的)方法.那就是我得到的.
I think this can be achieved somehow by a group by but I couldn't get it right so I choose another (sloppy) approach. Thats what i got.
data%>%
filter(Sensor=="10002") -> sensor_data_temp
sensor_data_temp$CaseID2 <- NA
case_id = 1
for(i in 1:nrow(sensor_data_temp)){
current_value <- sensor_data_temp[i,"values"]
next_value <- sensor_data_temp[i+1,"values"]
if(i+1 > nrow(sensor_data_temp)){
break
}
if(current_value==0 & next_value==1 || current_value==1 & next_value==1){
sensor_data_temp$CaseID2[i+1] <- case_id
}
else if(current_value==1 & next_value==0){
sensor_data_temp$CaseID2[i+1] <- 0
case_id = case_id +1
}
else{
sensor_data_temp$CaseID2[i+1] <- 0
}
}
我认为这就是我如何获取一个传感器的Caseid的方法.但是我不知道如何将每个传感器都放入一个数据帧中(如上一个)
I think thats how I could get the caseid's for one sensor. But I have no idea how I can manage to get every sensor into one dataframe (as the one above)
我敢肯定,有一种更优雅的方式来获得我想要的东西.
I am sure there is a much more elegant way to get what I want.
我希望有人能帮助我..在此先感谢!:)
I hope somebody can help me.. Thanks in advance! :)
推荐答案
这是一种方法:
library(dplyr)
mutate(group_by(arrange(data, sensor, time), sensor),
caseID = case_when(values != 0 ~ cumsum(diff(c(0, values)) > 0),
TRUE ~ 0L))
这篇关于在R中按每组变异的文章就介绍到这了,希望我们推荐的答案对大家有所帮助,也希望大家多多支持IT屋!