[英]Count the number of times each value appears in a row dataframe r
I have the following dataframe (79000 rows): 我有以下数据框(79000行):
ID P1 P2 P3 P4 P5 P6 P7 P8
1 38005 28002 38005 38005 28002 34002 NA NA
2 28002 28002 28002 38005 28002 NA NA NA
I want to count the number of times each number(code) appears in a row of dataframe. 我想计算每个数字(代码)出现在数据帧行中的次数。 So the ouput something like this:
所以输出是这样的:
38005 appears 3 28002 appears 2 34002 appears 1 NA appears 2
28002 appears 3 38005 appears 1 28002 appears 1 NA appears 3
So far I tried to find the most frequent number (code): 到目前为止,我试图找到最频繁的号码(代码):
df$frequency <-apply(df,1,function(x) names(which.max(table(x))))
But I don't know how to count the number of times each number(code) appears in a row. 但是我不知道如何计算每个数字(代码)连续出现的次数。
Using tidyverse
and reshape2
you can do: 使用
tidyverse
和reshape2
您可以执行以下操作:
df %>%
gather(var, val, -ID) %>% #Transforming the data from wide to long format
group_by(val, ID) %>% #Grouping
summarise(count = n()) %>% #Performing the count
dcast(ID~val, value.var = "count") #Reshaping the data
ID 28002 34002 38005 NA
1 1 2 1 3 2
2 2 4 NA 1 3
Showing the first two non-NA columns with the biggest count according ID: 显示前两个非NA列,其ID最多:
df %>%
gather(var, val, -ID) %>% #Transforming the data from wide to long format
group_by(val, ID) %>% #Grouping
mutate(temp = n()) %>% #Performing the count
group_by(ID) %>% #Grouping
mutate(temp2 = dense_rank(temp)) %>% #Creating the rank based on count
group_by(ID, val) %>% #Grouping
summarise(temp3 = first(temp2), #Summarising
temp = first(temp)) %>%
arrange(ID, desc(temp3)) %>% #Arranging
na.omit() %>% #Deleting the rows with NA
group_by(ID) %>%
mutate(temp4 = ifelse(temp3 == first(temp3) | temp3 == nth(temp3, 2), 1, 0)) %>% #Identifying the highest and the second highest count
filter(temp4 == 1) %>% #Selecting the highest and the second highest count
dcast(ID~val, value.var = "temp") #Reshaping the data
ID 28002 38005
1 1 2 3
2 2 4 1
ID <- c("P1","P2","P3","P4","P5","P6","P7","P8","P1","P2","P3","P4","P5","P6","P7","P8","P1")
count <-c("38005","28002","38005","38005","28002","34002","NA","NA","2","28002","28002","28002","38005","28002","NA","NA","NA")
df<- cbind.data.frame(ID,count)
table(df$count)
Use this code to find out the count 使用此代码找出计数
I think you're looking for this. 我想您正在寻找这个。
sort(table(unlist(df1[-1])), decreasing=TRUE)
# 31002 38005 24003 34002 28002
# 13222 13193 13019 13018 12625
This is, you're excluding column 1 that contains the IDs and "unlist" the rest of your data frame into a vector. 也就是说,您要排除包含ID的第1列,并将数据帧的其余部分“取消列出”到向量中。 The
table()
then counts the appearance of each value, which you also can sort()
. 然后,
table()
计算每个值的外观,您也可以对其进行sort()
。 Set option decreasing=TRUE
and the first two values are the two most frequent ones. 设置选项
decreasing=TRUE
,并且前两个值是两个最常使用的值。
If the output is getting to long because of a lot of values, you can include the code into a head(.)
. 如果由于很多值而导致输出变长,则可以将代码包含在
head(.)
。 The default length of the output is six, but you can limit it to two by specifying n=2
which gives you exactly what you want. 输出的默认长度为6,但是您可以通过指定
n=2
来将其限制为n=2
,这将为您提供所需的确切信息。 No need for any packages. 无需任何包装。
head(sort(table(unlist(df1[-1])), decreasing=TRUE), n=2)
# 31002 38005
# 13222 13193
DATA: 数据:
set.seed(42) # for sake of reproducibility
df1 <- data.frame(id=1:9750,
matrix(sample(c(38005, 28002, 34002, NA, 24003, 31002), 7.8e4,
replace=TRUE), nrow=9750,
dimnames=list(NULL, paste0("P", 1:8))))
data.table solution 数据表解决方案
#read sample data
dt <- fread( "ID P1 P2 P3 P4 P5 P6 P7 P8
1 38005 28002 38005 38005 28002 34002 NA NA
2 28002 28002 28002 38005 28002 NA NA NA")
#melt
dt.melt <- melt(dt, id = 1, measure = patterns("^P"), na.rm = FALSE)
#and cast
dcast( dt.melt, ID ~ value, fun = length, fill = 0 )
# ID 28002 34002 38005 NA
# 1: 1 2 1 3 2
# 2: 2 4 0 1 3
声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.