Moving Average of Previous Three Values in R

Moving average of previous three values in R

A simplier implementation of w_i_l_l's mavback function based on his mav function

mavback <- function(x,n){ filter(x, c(0, rep(1/n,n)), sides=1) }

How to calculate moving average from previous rows in data.table?

You could shift the values before applying frollmean with na.rm = TRUE argument:

df[order(store,week),desired:=frollmean(shift(demand),n=2,na.rm=T),by=.(store)][]

     store  week demand desired
    <char> <int>  <num>   <num>
 1:      A     3  19.18      NA
 2:      A     5     NA  19.180
 3:      A     6     NA  19.180
 4:      A     8  19.55     NaN
 5:      A     9  20.50  19.550
 6:      A    10     NA  20.025
 7:      B     1  20.75      NA
 8:      B     2  17.70  20.750
 9:      B     4  19.40  19.225
10:      B     7  17.52  18.550

Calculating moving average

Rolling Means/Maximums/Medians in the zoo package (rollmean)
MovingAverages in TTR
ma in forecast

Take average of last n values and move a value down at a time

Using zoo and dplyr:

library(dplyr)
library(zoo)

df <- df %>% mutate(moving_average = round(rollmeanr(col1, 7, fill = NA)))


> df
   col1 moving_average
1    96             NA
2    56             NA
3    53             NA
4    63             NA
5    70             NA
6    65             NA
7    72             68
8   111             70
9    65             71
10   58             72
11   59             71
12   74             72
13   57             71
14   70             71
15   89             67
16   60             67
17   60             67
18   52             66
19   65             65
20   58             65
21   84             67
22   54             62
23   55             61
24   51             60
25   72             63

How to calculate a moving average in R

This is a base R solution.

First get the dates as class date, then store the order of the months, define a range (span_month) and finally move over all windows.

Month_t <- as.Date( paste0(
             substr(Month,1,4),"-",substr(Month,5,length(Month)),"-1"),
             "%Y-%m-%d" )

Month_ord <- order(Month_t)

span_month=12

sapply( 1:((length(Month_ord) + 1) - span_month), 
        function(x) mean(emp[Month_ord[x]:Month_ord[x + (span_month - 1)]]) )
# [1] 13.00000 15.00000 17.00000 19.00000 21.00000 23.16667 25.16667 27.16667
# [9] 29.16667 31.16667 33.16667 35.16667 37.16667 39.00000 40.83333 42.66667
#[17] 44.33333 45.83333 47.50000

calculate moving average across columns in R

Assuming that the question intended c((9+13+8)/3, (14+2+10)/3, (16+7+1)/3)
as the values for 1962 and not the values shown there, rollmean can be used in either of the following manners. These one-liners give matrices as the result but as.data.frame can be used on the result if it is important that it be a data frame.

library(zoo)

t(apply(DF, 1, rollmean, 3))
##      1962   1963    1964    1965   1966
## 1 10.0000 8.3333  9.0000  6.6667 11.667
## 2  8.6667 6.0000 10.3333  9.3333 13.000
## 3  8.0000 8.6667  7.3333 10.0000  6.000

t(rollmean(t(DF), 3))
##      [,1]   [,2]    [,3]    [,4]   [,5]
## 1 10.0000 8.3333  9.0000  6.6667 11.667
## 2  8.6667 6.0000 10.3333  9.3333 13.000
## 3  8.0000 8.6667  7.3333 10.0000  6.000

Note

The input in reproducible form:

Lines <- "
   1961 1962 1963 1964 1965 1966 1967
1    9   13    8    4   15    1   19
2   14    2   10    6   15    7   17
3   16    7    1   18    3    9    6"
DF <- read.table(text = Lines, check.names = FALSE)

Using rollmean to calculate a moving average excluding the first observation in R

I found the simplest way to work this was with the lag function.

data$turnout <- lag(rollmean(data[,2], 1, fill = NA),1)

R - How to make a mean/average of n previous values, excluding current observation (rolling average)

We need to take the lag of the output derived from rollapply.

library(dplyr)
library(zoo)
LateCounts %>%
      mutate(RollAvge= lag(rollapplyr(Count, 12, mean, partial = TRUE)))

-output

#      Date Count RollAvge
#1  Jan-19     7       NA
#2  Feb-19     4 7.000000
#3  Mar-19     9 5.500000
#4  Apr-19     8 6.666667
#5  May-19     7 7.000000
#6  Jun-19     4 7.000000
#7  Jul-19     4 6.500000
#8  Aug-19     5 6.142857
#9  Sep-19     2 6.000000
#10 Oct-19     5 5.555556
#11 Nov-19     7 5.500000
#12 Dec-19     4 5.636364
#13 Jan-20     3 5.500000
#14 Feb-20     4 5.166667
#15 Mar-20     5 5.166667
#16 Apr-20     2 4.833333
#17 May-20     3 4.333333
#18 Jun-20     2 4.000000
#19 Jul-20     3 3.833333
#20 Aug-20     4 3.750000
#21 Sep-20     3 3.666667
#22 Oct-20     2 3.750000

data

LateCounts <- structure(list(Date = c("Jan-19", "Feb-19", "Mar-19", "Apr-19", 
"May-19", "Jun-19", "Jul-19", "Aug-19", "Sep-19", "Oct-19", "Nov-19", 
"Dec-19", "Jan-20", "Feb-20", "Mar-20", "Apr-20", "May-20", "Jun-20", 
"Jul-20", "Aug-20", "Sep-20", "Oct-20"), Count = c(7L, 4L, 9L, 
8L, 7L, 4L, 4L, 5L, 2L, 5L, 7L, 4L, 3L, 4L, 5L, 2L, 3L, 2L, 3L, 
4L, 3L, 2L)), class = "data.frame", row.names = c("1", "2", "3", 
"4", "5", "6", "7", "8", "9", "10", "11", "12", "13", "14", "15", 
"16", "17", "18", "19", "20", "21", "22"))

Moving Average of Previous Three Values in R