DF<-data.frame(id=c(1,2,3,4,5,6),date=c("2001-01-01","NA","2005-05-05","2006-06-06","2007-07-07","NA"))
id date
1 1 2001-01-01
2 2 NA
3 3 2005-05-05
4 4 2006-06-06
5 5 2007-07-07
6 6 NA
DF1<-data.frame(id=c(1,3,7,8,9,11),cdate=c("2000-12-20","2002-02-02","2003-03-03","2004-04-04","2005-05-05","2006-06-06"),atc=c("AAA1","BBB","AAA","DDD","CCC","AAA"))
id cdate atc
1 1 2000-12-20 AAA1
2 3 2002-02-02 BBB
3 7 2003-03-03 AAA
4 8 2004-04-04 DDD
5 9 2005-05-05 CCC
6 11 2006-06-06 AAA
DF2<-data.frame(id=c(1,2,3,4,8,9),cdate=c("2000-12-20","2002-02-02","2005-04-05","2004-04-04","2005-05-05","2006-06-06"),op=c("AA3","BB","AA","DD","CC","AA"))
id cdate op
1 1 2000-12-20 AA3
2 2 2002-02-02 BB
3 3 2005-04-05 AA
4 4 2004-04-04 DD
5 8 2005-05-05 CC
6 9 2006-06-06 AA
LM<-c("AAA","AA","BB")
Now I want to check if DF1 and DF2 contains any id in DF where atc/op starting with the values in LM. Finally, the cdate should not be more than 90 older than date.
Desired output:
id opatc
1 1 1
2 2 0
3 3 1
4 4 0
5 5 0
6 6 0
Which is the smartest way (shortest code) of solving this problem?
Best H
using tidyverse, you could do:
library(tidyverse)
DF %>%
left_join(DF1, "id")%>%
left_join(DF2, "id")%>%
mutate( cdate.x = as.Date(cdate.x),cdate.y = as.Date(cdate.y),
date = as.Date(date),
opatc = date - pmax(coalesce(cdate.x,cdate.y), cdate.y) <= 90) %>%
group_by(id)%>%
mutate(opatc = +(opatc&invoke("|",across(c(atc, op),
~any(sub("\\d+$","",.x)%in%LM)))))%>%
replace_na(list(opatc = 0))%>%
select(id, opatc)
# A tibble: 6 x 2
# Groups: id [6]
id opatc
<dbl> <dbl>
1 1 1
2 2 0
3 3 1
4 4 0
5 5 0
6 6 0
Related
I want to count the number of fluctuation of responses under the column response per id. However, responses No~ no ~ DK. I need to consider as the same response just only for count to the number of fluctuate in response. I don't change responses permanently.
df <- data.frame(
id=c(1,1,1,1,1,2,2,2,2,2,2,3,3,3,3,3,3,4,4,4,4,4,4),
response=c("Yes","Yes","No","DK","no","No","No","no","No","Yes","Yes","DK","No","Yes","Yes","No","No","No","died","TO","Yes","No","Yes")
)
I am trying it using the following code:
library(tidyverse)
df <- df %>%
group_by(id) %>% fill(response) %>%
mutate(new = rleid(response), rn = row_number()) %>%
mutate(flactuation = case_when(rn >2 & duplicated(new) ~ 'No', rn > 2 ~ 'Yes')) %>%
mutate(numberofchange = sum(flactuation=="Yes", na.rm = T)) %>% select(-rn, -flactuation)
Expected
id response new numberofchange
<dbl> <chr> <int> <int>
1 1 Yes 1 1
2 1 Yes 1 1
3 1 No 2 1
4 1 DK 2 1
5 1 no 2 1
6 2 No 1 1
7 2 No 1 1
8 2 no 1 1
9 2 No 1 1
10 2 Yes 2 1
11 2 Yes 2 1
12 3 DK 1 2
13 3 No 1 2
14 3 Yes 2 2
15 3 Yes 2 2
16 3 No 3 2
17 3 No 3 2
18 4 No 1 5
19 4 died 2 5
20 4 TO 3 5
21 4 Yes 4 5
22 4 No 5 5
23 4 Yes 6 5
You could use data.table::rleid() to get the run-length indices.
library(dplyr)
df %>%
group_by(id) %>%
mutate(new = data.table::rleid(replace(response, response %in% c('no', 'DK'), "No")),
numberofchange = max(new) - 1) %>%
ungroup()
# A tibble: 23 × 4
id response new numberofchange
<dbl> <chr> <int> <dbl>
1 1 Yes 1 1
2 1 Yes 1 1
3 1 No 2 1
4 1 DK 2 1
5 1 no 2 1
6 2 No 1 1
7 2 No 1 1
8 2 no 1 1
9 2 No 1 1
10 2 Yes 2 1
11 2 Yes 2 1
12 3 DK 1 2
13 3 No 1 2
14 3 Yes 2 2
15 3 Yes 2 2
16 3 No 3 2
17 3 No 3 2
18 4 No 1 5
19 4 died 2 5
20 4 TO 3 5
21 4 Yes 4 5
22 4 No 5 5
23 4 Yes 6 5
I have a series of rows in a single dataframe. I'm trying to aggregate the first two rows for each ID- i.e. - I want to combine events 1 and 2 for ID 1 into a single row, events 1 and 2 for ID 2 into a singlw row etc, but leave event 3 completely untouched.
id <- c(1,1,1,2,2,2,3,3,3,4,4,4,5,5,5)
event <- c(1,2,3,1,2,3,1,2,3,1,2,3,1,2,3)
score <- c(3,NA,1,3,NA,2,6,NA,1,8,NA,2,4,NA,1)
score2 <- c(NA,4,1,NA,5,2,NA,0,3,NA,5,6,NA,8,7)
df <- tibble(id, event, score, score2)
# A tibble: 15 x 4
id event score score2
<dbl> <dbl> <dbl> <dbl>
1 1 1 3 NA
2 1 2 NA 4
3 1 3 1 1
4 2 1 3 NA
5 2 2 NA 5
6 2 3 2 2
7 3 1 6 NA
8 3 2 NA 0
9 3 3 1 3
10 4 1 8 NA
11 4 2 NA 5
12 4 3 2 6
13 5 1 4 NA
14 5 2 NA 8
15 5 3 1 7
I've tried :
df_merged<- df %>% group_by (id) %>% summarise_all(funs(min(as.character(.),na.rm=TRUE))),
which aggregates these nicely, but then I struggle to merge these back into the orignal dataframe/tibble (there are really about 300 different "score" columns in the full dataset, so a right_join is a headache with score.x, score.y, score2.x, score2.y all over the place...)
Ideally, the situation would need to be dplyr as the rest of my code runs on this!
EDIT:
Ideally, my expected output would be:
# A tibble: 10 x 4
id event score score2
<dbl> <dbl> <dbl> <dbl>
1 1 1 3 4
3 1 3 1 1
4 2 1 3 5
6 2 3 2 2
7 3 1 6 0
9 3 3 1 3
10 4 1 8 5
12 4 3 2 6
13 5 1 4 8
15 5 3 1 7
We may change the order of NA elements with replace
library(dplyr)
df %>%
group_by(id) %>%
mutate(across(starts_with('score'),
~replace(., 1:2, .[1:2][order(is.na(.[1:2]))]))) %>%
ungroup %>%
filter(if_all(starts_with('score'), Negate(is.na)))
-output
# A tibble: 10 x 4
id event score score2
<dbl> <dbl> <dbl> <dbl>
1 1 1 3 4
2 1 3 1 1
3 2 1 3 5
4 2 3 2 2
5 3 1 6 0
6 3 3 1 3
7 4 1 8 5
8 4 3 2 6
9 5 1 4 8
10 5 3 1 7
Here is an alternative way to achieve your task with fill from tidyr package:
library(dplyr)
library(tidyr)
df %>%
group_by(id) %>%
fill(everything(), .direction = "down") %>%
fill(everything(), .direction = "up") %>%
slice(1,3)
id event score score2
<dbl> <dbl> <dbl> <dbl>
1 1 1 3 4
2 1 3 1 1
3 2 1 3 5
4 2 3 2 2
5 3 1 6 0
6 3 3 1 3
7 4 1 8 5
8 4 3 2 6
9 5 1 4 8
10 5 3 1 7
How about this?
library(dplyr)
df_e12 <- df %>%
filter(event %in% c(1, 2)) %>%
group_by(id) %>%
mutate(across(starts_with("score"), ~min(.x, na.rm = TRUE))) %>%
ungroup() %>%
distinct(id, .keep_all = TRUE)
df_e3 <- df %>%
filter(event == 3)
df <- bind_rows(df_e12, df_e3) %>%
arrange(id, event)
df
> df
# A tibble: 10 x 4
id event score score2
<dbl> <dbl> <dbl> <dbl>
1 1 1 3 4
2 1 3 1 1
3 2 1 3 5
4 2 3 2 2
5 3 1 6 0
6 3 3 1 3
7 4 1 8 5
8 4 3 2 6
9 5 1 4 8
10 5 3 1 7
DF<-data.frame(id=c(1,1,2,2,3,3),code=c("AA","BB","AA","DD","EE","FF"),date=c("2001-01-01","NA","2005-05-05","2006-06-06","2007-07-07","NA"),date2=c("2000-12-20","2000-12-20","2000-12-20","2000-12-20","2000-12-20","2000-12-20"))
DF
id code date date2
1 1 AA 2001-01-01 2000-12-20
2 1 BB NA 2000-12-20
3 2 AA 2005-05-05 2000-12-20
4 2 DD 2006-06-06 2000-12-20
5 3 EE 2007-07-07 2000-12-20
6 3 FF NA 2000-12-20
VAR=c("AA","HH","KK")
Desired output:
id code date date2 var
1 1 AA 2001-01-01 2000-12-20 1
2 1 BB NA 2000-12-20 1
3 2 AA 2005-05-05 2000-12-20 0
4 2 DD 2006-06-06 2000-12-20 0
5 3 EE 2007-07-07 2000-12-20 0
6 3 FF NA 2000-12-20 0
I want to check if any of the variables in VAR matches with code. If true then I want to check that date is max 30 days older than date2 - if that also is true it should result in 1 in var. Everything should ge grouped by id.
Thanks in advance
Best H
library(dplyr)
DF %>%
mutate_at(vars(date, date2), as.Date) %>%
group_by(id) %>%
mutate(var = +(any(code %in% VAR) & (is.na(date) | (date - date2) < 30))) %>%
ungroup()
# # A tibble: 6 x 5
# id code date date2 var
# <dbl> <chr> <date> <date> <int>
# 1 1 AA 2001-01-01 2000-12-20 1
# 2 1 BB NA 2000-12-20 1
# 3 2 AA 2005-05-05 2000-12-20 0
# 4 2 DD 2006-06-06 2000-12-20 0
# 5 3 EE 2007-07-07 2000-12-20 0
# 6 3 FF NA 2000-12-20 0
I have data in which subjects completed multiple ratings per day over 6-7 days. The number of ratings per day varies. The data set includes subject ID, date, and the ratings. I would like to create a new variable that recodes the dates for each subject into "study day" --- so 1 for first day of ratings, 2 for second day of ratings, etc.
For example, I would like to take this:
id Date Rating
1 10/20/2018 2
1 10/20/2018 3
1 10/20/2018 5
1 10/21/2018 1
1 10/21/2018 7
1 10/21/2018 9
1 10/22/2018 4
1 10/22/2018 5
1 10/22/2018 9
2 11/15/2018 1
2 11/15/2018 3
2 11/15/2018 4
2 11/16/2018 3
2 11/16/2018 1
2 11/17/2018 0
2 11/17/2018 2
2 11/17/2018 9
And end up with this:
id Day Date Rating
1 1 10/20/2018 2
1 1 10/20/2018 3
1 1 10/20/2018 5
1 2 10/21/2018 1
1 2 10/21/2018 7
1 2 10/21/2018 9
1 3 10/22/2018 4
1 3 10/22/2018 5
1 3 10/22/2018 9
2 1 11/15/2018 1
2 1 11/15/2018 3
2 1 11/15/2018 4
2 2 11/16/2018 3
2 2 11/16/2018 1
2 3 11/17/2018 0
2 3 11/17/2018 2
2 3 11/17/2018 9
I was going to look into setting up some kind of loop, but I thought it would be worth asking if there is a more efficient way to pull this off. Are there any functions that would allow me to automate this sort of thing? Thanks very much for any suggestions.
Since you want to reset the count after every id , makes this question a bit different.
Using only base R, we can split the Date based on id and then create a count of each distinct group.
df$Day <- unlist(sapply(split(df$Date, df$id), function(x) match(x,unique(x))))
df
# id Date Rating Day
#1 1 10/20/2018 2 1
#2 1 10/20/2018 3 1
#3 1 10/20/2018 5 1
#4 1 10/21/2018 1 2
#5 1 10/21/2018 7 2
#6 1 10/21/2018 9 2
#7 1 10/22/2018 4 3
#8 1 10/22/2018 5 3
#9 1 10/22/2018 9 3
#10 2 11/15/2018 1 1
#11 2 11/15/2018 3 1
#12 2 11/15/2018 4 1
#13 2 11/16/2018 3 2
#14 2 11/16/2018 1 2
#15 2 11/17/2018 0 3
#16 2 11/17/2018 2 3
#17 2 11/17/2018 9 3
I don't know how I missed this but thanks to #thelatemail who reminded that this is basically the same as
library(dplyr)
df %>%
group_by(id) %>%
mutate(Day = match(Date, unique(Date)))
AND
df$Day <- as.numeric(with(df, ave(Date, id, FUN = function(x) match(x, unique(x)))))
If you want a slightly hacky dplyr version....you can use the date column and convert it to a numeric date then manipulate that number to give the desired result
library(tidyverse)
library(lubridate)
df <- data_frame(id=c(1,1,1,1,1,1,1,1,1,2,2,2,2,2,2,2,2),
Date= c('10/20/2018', '10/20/2018', '10/20/2018', '10/21/2018', '10/21/2018', '10/21/2018',
'10/22/2018', '10/22/2018', '10/22/2018','11/15/2018', '11/15/2018', '11/15/2018',
'11/16/2018', '11/16/2018', '11/17/2018', '11/17/2018', '11/17/2018'),
Rating=c(2,3,5,1,7,9,4,5,9,1,3,4,3,1,0,2,9))
df %>%
group_by(id) %>%
mutate(
Date = mdy(Date),
Day = as.numeric(Date),
Day = Day-min(Day)+1)
# A tibble: 17 x 4
# Groups: id [2]
id Date Rating Day
<dbl> <date> <dbl> <dbl>
1 1 2018-10-20 2 1
2 1 2018-10-20 3 1
3 1 2018-10-20 5 1
4 1 2018-10-21 1 2
5 1 2018-10-21 7 2
6 1 2018-10-21 9 2
7 1 2018-10-22 4 3
8 1 2018-10-22 5 3
9 1 2018-10-22 9 3
10 2 2018-11-15 1 1
11 2 2018-11-15 3 1
12 2 2018-11-15 4 1
13 2 2018-11-16 3 2
14 2 2018-11-16 1 2
15 2 2018-11-17 0 3
16 2 2018-11-17 2 3
17 2 2018-11-17 9 3
I have an R dataframe such as:
df <- data.frame(period=rep(1:4,2),
farm=c(rep('A',4),rep('B',4)),
cumVol=c(1,5,15,31,10,12,16,24),
other = 1:8);
period farm cumVol other
1 1 A 1 1
2 2 A 5 2
3 3 A 15 3
4 4 A 31 4
5 1 B 10 5
6 2 B 12 6
7 3 B 16 7
8 4 B 24 8
How do I find the change in cumVol at each farm in each period, ignoring the 'other' column? I would like a dataframe like this (optionally with the cumVol column remaining):
period farm volume other
1 1 A 0 1
2 2 A 4 2
3 3 A 10 3
4 4 A 16 4
5 1 B 0 5
6 2 B 2 6
7 3 B 4 7
8 4 B 8 8
In practice there may be many 'farm'-like columns, and many 'other'-like (ie. ignored) columns. I'd like to be able to specify all the column names using variables.
I am using the dplyr package.
In dplyr:
require(dplyr)
df %>%
group_by(farm) %>%
mutate(volume = cumVol - lag(cumVol, default = cumVol[1]))
Source: local data frame [8 x 5]
Groups: farm
period farm cumVol other volume
1 1 A 1 1 0
2 2 A 5 2 4
3 3 A 15 3 10
4 4 A 31 4 16
5 1 B 10 5 0
6 2 B 12 6 2
7 3 B 16 7 4
8 4 B 24 8 8
Perhaps the desired output should actually be as follows?
df %>%
group_by(farm) %>%
mutate(volume = cumVol - lag(cumVol, default = 0))
period farm cumVol other volume
1 1 A 1 1 1
2 2 A 5 2 4
3 3 A 15 3 10
4 4 A 31 4 16
5 1 B 10 5 10
6 2 B 12 6 2
7 3 B 16 7 4
8 4 B 24 8 8
Edit: Following up on your comments I think you are looking for arrange(). It that is not the case it might be best to start a new question.
df1 <- data.frame(period=rep(1:4,4), farm=rep(c(rep('A',4),rep('B',4)),2), crop=(c(rep('apple',8), rep('pear',8))), cumCropVol=c(1,5,15,31,10,12,16,24,11,15,25,31,20,22,26,34), other = rep(1:8,2) );
df1 %>%
arrange(desc(period), desc(farm)) %>%
group_by(period, farm) %>%
summarise(cumVol=sum(cumCropVol))
Edit: Follow up #2
df1 <- data.frame(period=rep(1:4,4), farm=rep(c(rep('A',4),rep('B',4)),2), crop=(c(rep('apple',8), rep('pear',8))), cumCropVol=c(1,5,15,31,10,12,16,24,11,15,25,31,20,22,26,34), other = rep(1:8,2) );
df <- df1 %>%
arrange(desc(period), desc(farm)) %>%
group_by(period, farm) %>%
summarise(cumVol=sum(cumCropVol))
ungroup(df) %>%
arrange(farm) %>%
group_by(farm) %>%
mutate(volume = cumVol - lag(cumVol, default = 0))
Source: local data frame [8 x 4]
Groups: farm
period farm cumVol volume
1 1 A 12 12
2 2 A 20 8
3 3 A 40 20
4 4 A 62 22
5 1 B 30 30
6 2 B 34 4
7 3 B 42 8
8 4 B 58 16
In dplyr -- so you don't have to replace NAs
library(dplyr)
df %>%
group_by(farm)%>%
mutate(volume = c(0,diff(cumVol)))
period farm cumVol other volume
1 1 A 1 1 0
2 2 A 5 2 4
3 3 A 15 3 10
4 4 A 31 4 16
5 1 B 10 5 0
6 2 B 12 6 2
7 3 B 16 7 4
8 4 B 24 8 8
Would creating a new column in your original dataset be an option?
Here is an option using the data.table operator :=.
require("data.table")
DT <- data.table(df)
DT[, volume := c(0,diff(cumVol)), by="farm"]
or
diff_2 <- function(x) c(0,diff(x))
DT[, volume := diff_2(cumVol), by="farm"]
Output:
# > DT
# period farm cumVol other volume
# 1: 1 A 1 1 0
# 2: 2 A 5 2 4
# 3: 3 A 15 3 10
# 4: 4 A 31 4 16
# 5: 1 B 10 5 0
# 6: 2 B 12 6 2
# 7: 3 B 16 7 4
# 8: 4 B 24 8 8
tapply and transform?
> transform(df, volumen=unlist(tapply(cumVol, farm, function(x) c(0, diff(x)))))
period farm cumVol other volumen
A1 1 A 1 1 0
A2 2 A 5 2 4
A3 3 A 15 3 10
A4 4 A 31 4 16
B1 1 B 10 5 0
B2 2 B 12 6 2
B3 3 B 16 7 4
B4 4 B 24 8 8
ave is a better option, see # thelatemail's comment
with(df, ave(cumVol,farm,FUN=function(x) c(0,diff(x))) )