What are the pitfals of using kernel density estimates to infer about the shape of an underlying distribution

by Sebastian   Last Updated November 05, 2018 14:19 PM

I know that we cannot simply infer from the shape of a histogram to the shape of the underlying distribution, as the shape of the histogram is influenced by the choice of the intervals (Assessing approximate distribution of data based on a histogram)

However I wonder whether the use of the smoothed density estimate of ggplot2 also suffers from this problem to a similar extent? Furthermore what (other) problems or difficulties can arise when using kernel density estimates to e.g. decide whether a distribution is unimodal?

Related Questions

Benefit of smoothing survival function

Updated July 15, 2016 08:08 AM