Average percentage change over different years in R - r

I have a data frame from which I created a reproducible example:
country <- c('A','A','A','B','B','C','C','C','C')
year <- c(2010,2011,2015,2008,2009,2008,2009,2011,2015)
score <- c(1,2,2,1,4,1,1,3,2)
country year score
1 A 2010 1
2 A 2011 2
3 A 2015 2
4 B 2008 1
5 B 2009 4
6 C 2008 1
7 C 2009 1
8 C 2011 3
9 C 2015 2
And I am trying to calculate the average percentage increase (or decrease) in the score for each country by calculating [(final score - initial score) ÷ (initial score)] for each year and averaging it over the number of years.
country year score change
1 A 2010 1 NA
2 A 2011 2 1
3 A 2015 2 0
4 B 2008 1 NA
5 B 2009 4 3
6 C 2008 1 NA
7 C 2009 1 0
8 C 2011 3 2
9 C 2015 2 -0.33
The final result I am hoping to obtain:
country avg_change
1 A 0.5
2 B 3
3 C 0.55
As you can see, the trick is that countries have spans over different years, sometimes with a missing year in between. I tried different ways to do it manually but I do struggle. If someone could hint me a solution would be great. Many thanks.


With dplyr, we can group_by country and get mean of difference between scores.
library(dplyr)
df %>%
group_by(country) %>%
summarise(avg_change = mean(c(NA, diff(score)), na.rm = TRUE))
# country avg_change
# <fct> <dbl>
#1 A 0.500
#2 B 3.00
#3 C 0.333
Using base R aggregate with same logic
aggregate(score~country, df, function(x) mean(c(NA, diff(x)), na.rm = TRUE))


We can use data.table to group by 'country' and take the mean of the difference between the 'score' and the lag of 'score'
library(data.table)
setDT(df1)[, .(avg_change = mean(score -lag(score), na.rm = TRUE)), .(country)]
# country avg_change
#1: A 0.5000000
#2: B 3.0000000
#3: C 0.3333333

Related

Paste values in a column based on other observations in the dataframe in R

I have a very large (~30M observations) dataframe in R and I am having trouble with a new column I want to create.
The data is formatted like this:
Country Year Value
1 A 2000 1
2 A 2001 NA
3 A 2002 2
4 B 2000 4
5 B 2001 NA
6 B 2002 NA
7 B 2003 3
My problem is that I would like to impute the NAs in the value column based on other values in that column. Specifically, if there is a non-NA value for the same country I would like that to replace the NA in later years, until there is another non-NA value.
The data above would therefore be transformed into this:
Country Year Value
1 A 2000 1
2 A 2001 1
3 A 2002 2
4 B 2000 4
5 B 2001 4
6 B 2002 4
7 B 2003 3
To solve this, I first tried using a loop with a lookup function and also some if_else statements, but wasn't able to get it to behave as I expected. In general, I am struggling to find an efficient solution that will be able to perform the task in the order of minutes-hours and not days.
Is there an easy way to do this?
Thanks!

Using tidyr's fill:
library(tidyverse)
df %>%
group_by(Country) %>%
fill(Value)
Result:
# A tibble: 7 × 3
# Groups: Country [2]
Country Year Value
<chr> <dbl> <dbl>
1 A 2000 1
2 A 2001 1
3 A 2002 2
4 B 2000 4
5 B 2001 4
6 B 2002 4
7 B 2003 3

Create a new column with max values using the identifier column within a pipeline

I am trying to clean up some old code and convert over to "tidy". I am trying to create a new column of data within a pipeline that is the maximum age of individual fish. Let's represent the columns of interest as:
fish_1 <- data.frame(year = c(2012,2012,2015,2015,2015,2013,2013,2013,2013,2012,2012,2015,2015,2015),
fishid = c('a','a','b','b','b','c','c','c','c','d','d','e','e','e'), # unique identifier for each fish
agei = c(1,2,1,2,3,1,2,3,4,1,2,1,2,3))
# which looks like this:
fish_1
year fishid agei
1 2012 a 1
2 2012 a 2
3 2015 b 1
4 2015 b 2
5 2015 b 3
6 2013 c 1
7 2013 c 2
8 2013 c 3
9 2013 c 4
10 2012 d 1
11 2012 d 2
12 2015 e 1
13 2015 e 2
14 2015 e 3
What I'm trying to do is create a new column agec that is the maximum age for each individual fish repeated however many number of times is required to fill the rows for each fish.
The desired output would be:
fish_2 <- data.frame(year = c(2012,2012,2015,2015,2015,2013,2013,2013,2013,2012,2012,2015,2015,2015),
fishid = c('a','a','b','b','b','c','c','c','c','d','d','e','e','e'), # unique identifier for each fish
agei = c(1,2,1,2,3,1,2,3,4,1,2,1,2,3),
agec = c(2,2,3,3,3,4,4,4,4,2,2,3,3,3))
# Which looks like:
fish_2
year fishid agei agec
1 2012 a 1 2
2 2012 a 2 2
3 2015 b 1 3
4 2015 b 2 3
5 2015 b 3 3
6 2013 c 1 4
7 2013 c 2 4
8 2013 c 3 4
9 2013 c 4 4
10 2012 d 1 2
11 2012 d 2 2
12 2015 e 1 3
13 2015 e 2 3
14 2015 e 3 3
The way I had done this in the past was to use a plyr::ddply() call to create a new dataframe and then merge with fish like this:
caps = plyr::ddply(fish_1, c('fishid'), plyr::summarize, agec=max(agei))
fish = merge(fish_1, caps, by='fishid')
fish
fishid year agei agec
1 a 2012 1 2
2 a 2012 2 2
3 b 2015 1 3
4 b 2015 2 3
5 b 2015 3 3
6 c 2013 1 4
7 c 2013 2 4
8 c 2013 3 4
9 c 2013 4 4
10 d 2012 1 2
11 d 2012 2 2
12 e 2015 1 3
13 e 2015 2 3
14 e 2015 3 3
I'm hoping someone can help me achieve this data structure concisely within a pipeline. All of the similar questions I have found have been very verbose and not specific to this issue. I am new to using tidyverse but I'm having trouble getting the group_by() function (to replace the ddply() call) within a pipe, and I'm hoping there is a simpler way.
UPDATE
For those interested it appears both answers below are correct. The reason that I struggled was because I was already completing other data manipulations within my pipeline and I tried to complete the formation of the agec column within a previous call to dplyr::mutate(). You can refer to my comment on #Thomas answer to see the error in my ways. Hope this helps.

Try dplyr instead of plyr
library(dplyr)
fish_1 %>%
group_by(fishid) %>%
mutate(agec = max(agei))

You can use group_by from dplyr to group your fish IDs and then simply call mutate (dplyr as well) with max:
fish_1 <- data.frame(year = c(2012,2012,2015,2015,2015,2013,2013,2013,2013,2012,2012,2015,2015,2015),
fishid = c('a','a','b','b','b','c','c','c','c','d','d','e','e','e'), # unique identifier for each fish
agei = c(1,2,1,2,3,1,2,3,4,1,2,1,2,3))
fish_1 %>%
group_by(fishid) %>%
mutate(agec = max(agei))
# A tibble: 14 x 4
# Groups: fishid [5]
year fishid agei agec
<dbl> <chr> <dbl> <dbl>
1 2012 a 1 2
2 2012 a 2 2
3 2015 b 1 3
4 2015 b 2 3
5 2015 b 3 3
6 2013 c 1 4
7 2013 c 2 4
8 2013 c 3 4
9 2013 c 4 4
10 2012 d 1 2
11 2012 d 2 2
12 2015 e 1 3
13 2015 e 2 3
14 2015 e 3 3

An option with data.table
library(data.table)
setDT(fish_1)[, agec := max(agei, na.rm = TRUE), fishid]

Calculate average of values in R and add result as new rows instead of as a new column

I have a dataframe like the following one:
day year value
1 2014 5
1 2015 16
1 2016 0
2 2014 3
2 2015 1
2 2016 4
and I want to calculate the average value by day for the three year period (2014, 2015, 2016). The following code works for this purpose:
data %>%
group_by(day) %>%
mutate(MEAN = mean(value))
and produces this output:
day year value MEAN
1 2014 5 7
1 2015 16 7
1 2016 0 7
2 2014 3 3
2 2015 1 3
2 2016 4 3
but I want to add the average values as new rows in the same dataframe as follows:
day year value
1 2014 5
1 2015 16
1 2016 0
2 2014 3
2 2015 1
2 2016 4
1 avg 7 <--
2 avg 3 <--
Any suggestions about how can I possibly do this? Thanks!

We can use summarise (instead of mutate - which adds a new column in the original dataset) to calculate the mean and then with bind_rows can bind with original data. The tidyverse functions are very particular about type, so make sure the class are the same before we do the binding
library(dplyr)
data %>%
group_by(day) %>%
summarise(year = 'avg', value = mean(value)) %>%
bind_rows(data %>%
mutate(year = as.character(year)), .)
# day year value
#1 1 2014 5.00
#2 1 2015 16.00
#3 1 2016 0.00
#4 2 2014 3.00
#5 2 2015 1.00
#6 2 2016 4.00
#7 1 avg 7.00
#8 2 avg 2.67
Another option is to split by the 'day' and then with add_row (from tibble) create a new row on each of the list elements
library(tibble)
library(purrr)
data %>%
mutate(year = as.character(year)) %>%
group_split(day) %>%
map_dfr(~ .x %>% add_row(day = first(.$day),
year = 'avg', value = mean(.$value)))

Here is a base R option using aggregate
rbind(df,cbind(aggregate(value~day,df,mean),year = "avg")[c(1,3,2)])
or a variation (by #thelatemail from comments)
rbind(df, aggregate(df["value"], cbind(df["day"], year="avg"), FUN=mean))
which gives
day year value
1 1 2014 5.000000
2 1 2015 16.000000
3 1 2016 0.000000
4 2 2014 3.000000
5 2 2015 1.000000
6 2 2016 4.000000
7 1 avg 7.000000
8 2 avg 2.666667

r conditional subtract number

I am trying to do the following logic to create 'subtract' column.
I have years from 1986-2014 and around 100 firms.
year firm count sum_of_year subtract
1986 A 1 2 2
1986 B 1 2 4
1987 A 2 4 5
1987 C 1 4 2
1987 D 1 4 5
1988 C 3 5
1988 E 2 5
That is, if a firm i at t appears in t+1, then subtract its count at t+1 from the sum_of_year at t+1,
if a firm i does not appear in t+1, then just put sum_of_year at t+1 as shown in the sample.
I am having difficulties in creating this conditional code.
How can I do this in a generalized version?
Thank you for your help.

One way using dplyr with the help of tidyr::complete. We complete the missing combinations of rows for year and firm and fill count with 0. For each year, we subtract the count by sum of count for that entire year and finally for each firm, we take the value from the next year using lead.
library(dplyr)
df %>%
tidyr::complete(year, firm, fill = list(count = 0)) %>%
group_by(year) %>%
mutate(n = sum(count) - count) %>%
group_by(firm) %>%
mutate(subtract = lead(n)) %>%
filter(count != 0) %>%
select(-n)
# year firm count sum_of_year subtract
# <int> <fct> <dbl> <int> <dbl>
#1 1986 A 1 2 2
#2 1986 B 1 2 4
#3 1987 A 2 4 5
#4 1987 C 1 4 2
#5 1987 D 1 4 5
#6 1988 C 3 5 NA
#7 1988 E 2 5 NA

expand.grid() based on values in two variables in R

I would like to expand a grid in R such that the expansion occurs for unique values of one variable but joint values for two variables. For example:
frame <- data.frame(id = seq(1:2),id2 = seq(1:2), year = c(2005, 2008))
I would like to expand the frame for each year, but such that id and id2 are considered jointly (e.g. (1,1), and (2,2) to generate an output like:
id id2 year
1 1 2005
1 1 2006
1 1 2007
1 1 2005
2 2 2006
2 2 2007
2 2 2008
Using expand.grid(), does someone know how to do this? I have not been able to wrangle the code past looking at each id uniquely and producing a frame with all combinations given the following code:
with(frame, expand.grid(year = seq(min(year), max(year)), id = unique(id), id2 = unique(id2)))
Thanks for any and all help.

You could do this with reshape::expand.grid.df
require(reshape)
expand.grid.df(data.frame(id=1:2,id2=1:2), data.frame(year=c(2005:2008)))
> expand.grid.df(data.frame(id=1:2,id2=1:2), data.frame(year=c(2005:2008)))
id id2 year
1 1 1 2005
2 2 2 2005
3 1 1 2006
4 2 2 2006
5 1 1 2007
6 2 2 2007
7 1 1 2008
8 2 2 2008

Here is another way using base R
indx <- diff(frame$year)+1
indx1 <- rep(1:nrow(frame), each=indx)
frame1 <- transform(frame[indx1,1:2], year=seq(frame$year[1], length.out=indx, by=1))
row.names(frame1) <- NULL
frame1
# id id2 year
#1 1 1 2005
#2 1 1 2006
#3 1 1 2007
#4 1 1 2008
#5 2 2 2005
#6 2 2 2006
#7 2 2 2007
#8 2 2 2008

Resources