# How to simulate the different types of missing data

How do you create a missingness mechanism (MAR, MCAR, NMAR)? Can you generate it directly or do you do it by a model?

• Welcome to CV! Software questions are generally off-topic here, and your question as it stands is rather broad and unclear. Could you edit to be more specific? Dec 2, 2015 at 22:49
• You mean that you have a complete dataset and you want to create one of these mechanisms , anyway I do not think you need R package ,if you want to create for example MNAR , set the variable value to NA conditional on its own value , for MAR set the variable value to NA conditional on the another variable value , for MCAR sample randomly from your dataset and set the sampled values to NA . Dec 2, 2015 at 23:12
• yes exactly thanks for your explain but can you give me an example please . Dec 2, 2015 at 23:19
• @user96940 suppose you have a complete dataset with 5 observations and two variables $X1$ and $X2$ (two columns ) , $X1$ realizations are (3,5,3,4,5) , $X2$ realizations are (55,60,55,65,55) ,to create MNAR on $X2$ ,you have to consider the realizations of $X2$ missing if e.g $X2$=55 , so $X2$ vector will be (NA,60,NA,65,NA) ,for MAR consider the realizations of $X2$ missing if $X1$=5 then $X2$ will be (55,NA,55,65,NA) for MCAR sample observations randomly from your dataset and set $X2$=NA for the sampled observations . Dec 3, 2015 at 1:13
• @BahgatNassour that is very useful , thanks a lot for your explain . Dec 3, 2015 at 1:28

Rubin defined three types of missing data:

1. Missing Completely at Random (MCAR)

MCAR occurs when there is a simple probability that data will be missing, and that probability is unrelated to anything else in your study. For example, a patient can miss a follow up visit because there is an accident on the highway and they simply can't get to the visit.

2. Missing at Random (MAR)

MAR happens when the missingness is related to information in your study, but all the relevant information to predict missingness is in the existing dataset. An example might be a weight loss study in which people drop out if their trajectory is that they are gaining weight. If you can estimate that trajectory for each person before anyone drops out, and see that those whose slope is positive subsequently drop out, you could take that as MAR.

3. Not Missing at Random (NMAR)

NMAR is like MAR in that the missingness is related to what is happening in your study, but differs in that the data that are related to the missingness is included in the data that are missing. For instance, if you are studying a treatment for vertigo / 'woozy-ness', but anytime a patient is really woozy, they don't show up for the follow-up visit. Thus, all the high values are missing, and they are missing because they are high!

In other words, the types of missingness specify the mechanism that generates the missingness itself, so if you understand how the mechanism works, you simply write code to replicate it. For example, if you want 7% of your data missing completely at random, draw a number from a uniform distribution for every value in your dataset, and if it is <.07, replace the value with NA. For missing at random, simulate a logistic regression data generating process that outputs a probability of each value being missing (i.e., being replaced with NA) using information that will continue to be non-missing in your dataset. (For an example of simulating a logistic regression data generating process, see my answer here: Logistic regression simulation in order to show that intercept is biased when Y=1 is rare.) You can generate missingness not at random using a similar logistic regression data generating process, where the probability of missingness is a function of the y-value itself (i.e., the value that will potentially be replaced by NA).

Here is an example:

##### generic data setup:
set.seed(977) # this makes the simulation exactly reproducible
ni     = 100  # 100 people
nj     =  10  # 10 week study
id     = rep(1:ni, each=nj)
cond   = rep(c("control", "diet"), each=nj*(ni/2))
base   = round(rep(rnorm(ni, mean=250, sd=10), each=nj))
week   = rep(1:nj, times=ni)
y      = round(base + rnorm(ni*nj, mean=0, sd=1))

# MCAR
prop.m = .07  # 7% missingness
mcar   = runif(ni*nj, min=0, max=1)
y.mcar = ifelse(mcar<prop.m, NA, y)  # unrelated to anything
View(cbind(id, week, cond, base, y, y.mcar))

# MAR
y.mar = matrix(y, ncol=nj, nrow=ni, byrow=TRUE)
for(i in 1:ni){
for(j in 4:nj){
dif1 = y.mar[i,j-2]-y.mar[i,j-3]
dif2 = y.mar[i,j-1]-y.mar[i,j-2]
if(dif1>0 & dif2>0){  # if weight goes up twice, drops out
y.mar[i,j:nj] = NA;  break
}
}
}
y.mar = as.vector(t(y.mar))
View(cbind(id, week, cond, base, y, y.mar))

# NMAR
sort.y = sort(y, decreasing=TRUE)
nmar   = sort.y[ceiling(prop.m*length(y))]
y.nmar = ifelse(y>nmar, NA, y)  # doesn't show up when heavier
View(cbind(id, week, cond, base, y, y.nmar))

• thanks for your explain but if may data is wine data set from UCI(university California Irvine ), How I simulate 5% ,10%, 15% missing ratios under three different modalities namely (MCAR ,<MAR ,NMAR ) for wine dataset . Dec 3, 2015 at 23:02
• wine data set dosnot contain any miss value, I want artificially generate missing values under (MCAR,MAR ,NMAR) . Dec 3, 2015 at 23:06
• @user96940, you do it just a I showed. MCAR would be exactly as I showed; for MAR & NMAR you have to decide what you want to be missing & why, then you'd code that up analogously to my example. Dec 4, 2015 at 1:01
• but I want all variables(attributes) from wine data contain missing value not just my class in your example you put NA just in variable class (y) , can you illustrate it by using my (wine dataset) if you can ?. thanks allot for help . Dec 4, 2015 at 13:41
• Just do the same thing for all variables. You would do MCAR exactly the same as I showed. NMAR would be analogous. MAR might be more difficult, but you would use the same type of reasoning as I have already demonstrated. Dec 4, 2015 at 13:54