I have a sample (N=2000). I have made a density plot to visualize the time trend for getting diagnosed with a disease, by using R, and specifically I have used ggplot. I have visualised the density plot by sex (one plot with two curves).
There are less than 4 individuals at some time events (I know this by counting). These time events that reveal less than 4 individuals, is shown through the time, such as at the beginning, in the middle and also at the end. Overall the density plot is showing very valuable information about the time trend for the disease.
However, I’m not allowed to report data for observations less than 4. I have tried to remove the data with less than 4 observations at the specific time events, but that made the density plot less pretty.
Can I use arguments such as adjust
or bw
in geom_density
to blur the data? In other words, to make it impossible to calculate the percentages of the total individuals for a specific time point?
I have thought about a histogram, but that will not show a clear trend.
So how can I make a plot in R, and still preserving the privacy of the data?