mydata <-data.frame(group = c(1,1,1,1,1,2,2,2,2,2), score = c(10, NA, NA, 20, 30, 5, NA, NA, NA, 40))
From 'mydata' I am trying to use dplyr to interpolate 'x' using na.approx when there are fewer than 3 consecutive NAs between the closest non-NA entries in 'value'. The interpolated x values are store in 'x_approx'.
Without the condition on the number of consecutive NAs in 'value' I use this code:
library(zoo)
mydata %>%
group_by(group) %>%
mutate(score_approx = na.approx(score)) %>%
mutate(score_approx = coalesce(score_approx,score))
mydata
# A tibble: 10 x 3
# Groups: group [2]
group score score_approx
<dbl> <dbl> <dbl>
1 1 10 10
2 1 NA 13.3
3 1 NA 16.7
4 1 20 20
5 1 30 30
6 2 5 5
7 2 NA 13.8
8 2 NA 22.5
9 2 NA 31.2
10 2 40 40
However, the desired data frame is:
# A tibble: 10 x 3
# Groups: group [2]
group score score_approx
<dbl> <dbl> <dbl>
1 1 10 10
2 1 NA 13.3
3 1 NA 16.7
4 1 20 20
5 1 30 30
6 2 5 5
7 2 NA NA
8 2 NA NA
9 2 NA NA
10 2 40 40
You can use maxgap argument in na.approx -
library(dplyr)
library(zoo)
mydata %>%
group_by(group) %>%
mutate(score_approx = na.approx(score, maxgap = 2)) %>%
ungroup
# group score score_approx
# <dbl> <dbl> <dbl>
# 1 1 10 10
# 2 1 NA 13.3
# 3 1 NA 16.7
# 4 1 20 20
# 5 1 30 30
# 6 2 5 5
# 7 2 NA NA
# 8 2 NA NA
# 9 2 NA NA
#10 2 40 40
Related
I have a dataframe like this:
df <- data_frame(id = c(rep('A', 10), rep('B', 10)),
value = c(1:3, rep(NA, 2), 1:2, rep(NA, 3), 1, rep(NA, 4), 1:3, rep(NA, 2)))
I need to count the number of consective NA's in the value column. The count needs to be grouped by ID, and it needs to restart at 1 every time a new NA or new series of NA's is encountered. The exptected output should look like this:
df$expected_output <- c(rep(NA, 3), 1:2, rep(NA, 2), 1:3, NA, 1:4, rep(NA, 3), 1:2)
If anyone can give me a dplyr solution that would also be great :)
I've tried a few things but nothing is giving any sort of sensical result. Thanks in advance^!
A solution using dplyr and data.table.
library(dplyr)
library(data.table)
df2 <- df %>%
group_by(id) %>%
mutate(info = rleid(value)) %>%
group_by(id, info) %>%
mutate(expected_output = row_number()) %>%
ungroup() %>%
mutate(expected_output = ifelse(!is.na(value), NA, expected_output)) %>%
select(-info)
df2
# # A tibble: 20 x 3
# id value expected_output
# <chr> <dbl> <int>
# 1 A 1 NA
# 2 A 2 NA
# 3 A 3 NA
# 4 A NA 1
# 5 A NA 2
# 6 A 1 NA
# 7 A 2 NA
# 8 A NA 1
# 9 A NA 2
# 10 A NA 3
# 11 B 1 NA
# 12 B NA 1
# 13 B NA 2
# 14 B NA 3
# 15 B NA 4
# 16 B 1 NA
# 17 B 2 NA
# 18 B 3 NA
# 19 B NA 1
# 20 B NA 2
We can use rle to get length of groups that are or are not na, and use purrr::map2 to apply seq if they are NA and get the growing count or just fill in with NA values using rep.
library(tidyverse)
count_na <- function(x) {
r <- rle(is.na(x))
consec <- map2(r$lengths, r$values, ~ if (.y) seq(.x) else rep(NA, .x))
unlist(consec)
}
df %>%
mutate(expected_output = count_na(value))
#> # A tibble: 20 × 3
#> id value expected_output
#> <chr> <dbl> <int>
#> 1 A 1 NA
#> 2 A 2 NA
#> 3 A 3 NA
#> 4 A NA 1
#> 5 A NA 2
#> 6 A 1 NA
#> 7 A 2 NA
#> 8 A NA 1
#> 9 A NA 2
#> 10 A NA 3
#> 11 B 1 NA
#> 12 B NA 1
#> 13 B NA 2
#> 14 B NA 3
#> 15 B NA 4
#> 16 B 1 NA
#> 17 B 2 NA
#> 18 B 3 NA
#> 19 B NA 1
#> 20 B NA 2
Here is a solution using rle:
x <- rle(is.na(df$value))
df$new[is.na(df$value)] <- sequence(x$lengths[x$values])
# A tibble: 20 x 3
id value new
<chr> <dbl> <int>
1 A 1 NA
2 A 2 NA
3 A 3 NA
4 A NA 1
5 A NA 2
6 A 1 NA
7 A 2 NA
8 A NA 1
9 A NA 2
10 A NA 3
11 B 1 NA
12 B NA 1
13 B NA 2
14 B NA 3
15 B NA 4
16 B 1 NA
17 B 2 NA
18 B 3 NA
19 B NA 1
20 B NA 2
Yet another solution:
library(tidyverse)
df %>%
mutate(aux =data.table::rleid(value)) %>%
group_by(id, aux) %>%
mutate(eout = ifelse(is.na(value), row_number(), NA_real_)) %>%
ungroup %>% select(-aux)
#> # A tibble: 20 × 4
#> id value expected_output eout
#> <chr> <dbl> <int> <dbl>
#> 1 A 1 NA NA
#> 2 A 2 NA NA
#> 3 A 3 NA NA
#> 4 A NA 1 1
#> 5 A NA 2 2
#> 6 A 1 NA NA
#> 7 A 2 NA NA
#> 8 A NA 1 1
#> 9 A NA 2 2
#> 10 A NA 3 3
#> 11 B 1 NA NA
#> 12 B NA 1 1
#> 13 B NA 2 2
#> 14 B NA 3 3
#> 15 B NA 4 4
#> 16 B 1 NA NA
#> 17 B 2 NA NA
#> 18 B 3 NA NA
#> 19 B NA 1 1
#> 20 B NA 2 2
I want to group my data in different chunks when the data is continuous. Trying to get the group column from dummy data like this:
a b group
<dbl> <dbl> <dbl>
1 1 1 1
2 2 2 1
3 3 3 1
4 4 NA NA
5 5 NA NA
6 6 NA NA
7 7 12 2
8 8 15 2
9 9 NA NA
10 10 25 3
I tried using
test %>% mutate(test = complete.cases(.)) %>%
group_by(group = cumsum(test == TRUE)) %>%
select(group, everything())
But it doesn't work as expected:
group a b test
<int> <dbl> <dbl> <lgl>
1 1 1 1 TRUE
2 2 2 2 TRUE
3 3 3 3 TRUE
4 3 4 NA FALSE
5 3 5 NA FALSE
6 3 6 NA FALSE
7 4 7 12 TRUE
8 5 8 15 TRUE
9 5 9 NA FALSE
10 6 10 25 TRUE
Any advice?
Using rle in base R -
transform(df, group1 = with(rle(!is.na(b)), rep(cumsum(values), lengths))) |>
transform(group1 = replace(group1, is.na(b), NA))
# a b group group1
#1 1 1 1 1
#2 2 2 1 1
#3 3 3 1 1
#4 4 NA NA NA
#5 5 NA NA NA
#6 6 NA NA NA
#7 7 12 2 2
#8 8 15 2 2
#9 9 NA NA NA
#10 10 25 3 3
A couple of approaches to consider if you wish to use dplyr for this.
First, you could look at transition from non-complete cases (using lag) to complete cases.
library(dplyr)
test %>%
mutate(test = complete.cases(.)) %>%
group_by(group = cumsum(test & !lag(test, default = F))) %>%
mutate(group = replace(group, !test, NA))
Alternatively, you could add row numbers to your data.frame. Then, you could filter to include only complete cases, and group_by enumerating with cumsum based on gaps in row numbers. Then, join back to original data.
test$rn <- seq.int(nrow(test))
test %>%
filter(complete.cases(.)) %>%
group_by(group = c(0, cumsum(diff(rn) > 1)) + 1) %>%
right_join(test) %>%
arrange(rn) %>%
dplyr::select(-rn)
Output
a b group
<int> <int> <dbl>
1 1 1 1
2 2 2 1
3 3 3 1
4 4 NA NA
5 5 NA NA
6 6 NA NA
7 7 12 2
8 8 15 2
9 9 NA NA
10 10 25 3
Using data.table, get rleid then remove group IDs for NAs, then fix the sequence with factor to integer conversion:
library(data.table)
setDT(test)[, group1 := {
x <- complete.cases(test)
grp <- rleid(x)
grp[ !x ] <- NA
as.integer(factor(grp))
}]
# a b group group1
# 1: 1 1 1 1
# 2: 2 2 1 1
# 3: 3 3 1 1
# 4: 4 NA NA NA
# 5: 5 NA NA NA
# 6: 6 NA NA NA
# 7: 7 12 2 2
# 8: 8 15 2 2
# 9: 9 NA NA NA
# 10: 10 25 3 3
For each id, I am trying to replace missing values with data that is available.
library(tidyverse)
df <- data.frame(id=c(1,1,1,2,2,2,3),
a=c(NA, NA, 10, NA, 12, NA, 10),
b=c(10, NA, NA, NA, 13,NA, NA))
> df
id a b
1 1 NA 10
2 1 NA NA
3 1 10 NA
4 2 NA NA
5 2 12 13
6 2 NA NA
7 3 10 NA
I have tried:
df %>%
dplyr::group_by(id) %>%
dplyr::mutate_at(vars(a:b), fill(., direction="up"))
and get the following error:
Error: 1 components of `...` had unexpected names.
We detected these problematic arguments:
* `direction`
Did you misspecify an argument?
Desired output:
id a b
1 1 10 10
2 1 10 NA
3 1 10 NA
4 2 12 13
5 2 12 13
6 2 12 13
7 3 10 NA
We dont' use fill with mutate_at. According to ?fill
data - A data frame. and
... - A selection of columns. If empty, nothing happens. You can supply bare variable names, select all variables between x and z with x:z, exclude y with -y. F
library(dplyr)
library(tidyr)
df %>%
group_by(id) %>%
fill(a:b, .direction = 'up')
# A tibble: 7 x 3
# Groups: id [3]
# id a b
# <dbl> <dbl> <dbl>
#1 1 10 10
#2 1 10 NA
#3 1 10 NA
#4 2 12 13
#5 2 12 13
#6 2 NA NA
#7 3 10 NA
I have a long format dataset with longitudinal data and for one variable I want to fill in the missings in timepoint 0 with the values in timepoint 1, but I do not want to fill in the missings from timepoint 1 with values from timepoint 2 and so on.
My dataset is ordered by id and timepoint.
I have used the fill function succesfully in cases where I just needed to fill missings from all timepoints from a specific id.
Example dataframe:
df <- data.frame(id=c(1,1,1,1,2,2,2,2,3,3,3,3,4,4,4,4),
timepoint=c(0,1,2,3,0,1,2,3,0,1,2,3,0,1,2,3),
var1=c(NA,9,8,10, NA, 10, NA, 12, NA, NA, 12, 11, NA, 12, 12, NA))
> df
id timepoint var1
1 1 0 NA
2 1 1 9
3 1 2 8
4 1 3 10
5 2 0 NA
6 2 1 10
7 2 2 NA
8 2 3 12
9 3 0 NA
10 3 1 NA
11 3 2 12
12 3 3 11
13 4 0 NA
14 4 1 12
15 4 2 12
16 4 3 NA
This is what works when I just need to fill any missing no matter the timepoint:
library(dplyr)
library(tidyr)
df <- df %>%
group_by(id) %>%
fill(`var9`:`var12`, .direction = "up") %>%
as.data.frame
But now I have trouble specifying to only fill in the missings in rows at timepoint 0. Any help is appreciated.
My expected output:
> df
id timepoint var1
1 1 0 9
2 1 1 9
3 1 2 8
4 1 3 10
5 2 0 10
6 2 1 10
7 2 2 NA
8 2 3 12
9 3 0 NA
10 3 1 NA
11 3 2 12
12 3 3 11
13 4 0 12
14 4 1 12
15 4 2 12
16 4 3 NA
This might be an oversimplification, but you can just call the fill function again, but this time with direction down. Then your entire data frame will be complete.
df <- data.frame(id=c(1,1,1,1,2,2,2,2,3,3,3,3,4,4,4,4),
timepoint=c(0,1,2,3,0,1,2,3,0,1,2,3,0,1,2,3),
var1=c(NA,9,8,10, NA, 10, NA, 12, NA, NA, 12, 11, NA, 12, 12, NA))
In this case I will use an ifelse statement followed the by the lead function.
library(dplyr); library(tidyr);
df %>%
group_by(id) %>%
mutate(var1 = ifelse(is.na(var1) & timepoint == 0,
lead(var1, 1), var1))
Yields:
# A tibble: 16 x 3
# Groups: id [4]
id timepoint var1
<dbl> <dbl> <dbl>
1 1 0 9
2 1 1 9
3 1 2 8
4 1 3 10
5 2 0 10
6 2 1 10
7 2 2 NA
8 2 3 12
9 3 0 NA
10 3 1 NA
11 3 2 12
12 3 3 11
13 4 0 12
14 4 1 12
15 4 2 12
16 4 3 NA
We can group_by id and use replace to change the values where timepoint = 0 & var1 is NA from the corresponding value of var1 where timepoint = 1 in each group.
library(dplyr)
df %>%
group_by(id) %>%
mutate(var2 = replace(var1, timepoint == 0 & is.na(var1), var1[timepoint == 1]))
# id timepoint var1 var2
# <dbl> <dbl> <dbl> <dbl>
# 1 1 0 NA 9
# 2 1 1 9 9
# 3 1 2 8 8
# 4 1 3 10 10
# 5 2 0 NA 10
# 6 2 1 10 10
# 7 2 2 NA NA
# 8 2 3 12 12
# 9 3 0 NA NA
#10 3 1 NA NA
#11 3 2 12 12
#12 3 3 11 11
#13 4 0 NA 12
#14 4 1 12 12
#15 4 2 12 12
#16 4 3 NA NA
set.seed(123)
df <- data.frame(x = sample(1:10, 20, replace = T), id = rep(1:2, each = 10))
For each id, I want to create a column which has the sum of previous 5 x values.
df %>% group_by(id) %>% mutate(roll.sum = c(x[1:4], zoo::rollapply(x, 5, sum)))
# Groups: id [2]
x id roll.sum
<int> <int> <int>
3 1 3
8 1 8
5 1 5
9 1 9
10 1 10
1 1 36
6 1 39
9 1 40
6 1 41
5 1 37
10 2 10
5 2 5
7 2 7
6 2 6
2 2 2
9 2 39
3 2 32
1 2 28
4 2 25
10 2 29
The 6th row should be 35 (3 + 8 + 5 + 9 + 10), the 7th row should be 33 (8 + 5 + 9 + 10 + 1) and so on.
However, the above function is also including the row itself for calculation. How can I fix it?
library(zoo)
df %>% group_by(id) %>%
mutate(Sum_prev = rollapply(x, list(-(1:5)), sum, fill=NA, align = "right", partial=F))
#you can use rollapply(x, list((1:5)), sum, fill=NA, align = "left", partial=F)
#to sum the next 5 elements scaping the current one
x id Sum_prev
1 3 1 NA
2 8 1 NA
3 5 1 NA
4 9 1 NA
5 10 1 NA
6 1 1 35
7 6 1 33
8 9 1 31
9 6 1 35
10 5 1 32
11 10 2 NA
12 5 2 NA
13 7 2 NA
14 6 2 NA
15 2 2 NA
16 9 2 30
17 3 2 29
18 1 2 27
19 4 2 21
20 10 2 19
There is the rollify function in the tibbletime package that you could use. You can read about it in this vignette: Rolling calculations in tibbletime.
library(tibbletime)
library(dplyr)
rollig_sum <- rollify(.f = sum, window = 5)
df %>%
group_by(id) %>%
mutate(roll.sum = lag(rollig_sum(x))) #added lag() here
# A tibble: 20 x 3
# Groups: id [2]
# x id roll.sum
# <int> <int> <int>
# 1 3 1 NA
# 2 8 1 NA
# 3 5 1 NA
# 4 9 1 NA
# 5 10 1 NA
# 6 1 1 35
# 7 6 1 33
# 8 9 1 31
# 9 6 1 35
#10 5 1 32
#11 10 2 NA
#12 5 2 NA
#13 7 2 NA
#14 6 2 NA
#15 2 2 NA
#16 9 2 30
#17 3 2 29
#18 1 2 27
#19 4 2 21
#20 10 2 19
If you want the NAs to be some other value, you can use, for example, if_else
df %>%
group_by(id) %>%
mutate(roll.sum = lag(rollig_sum(x))) %>%
mutate(roll.sum = if_else(is.na(roll.sum), x, roll.sum))