Kurtosis and Skewness
• kurt(A, B, C, ...)—Returns the kurtosis of the elements A, B, C, .... The kurtosis is defined as follows:
The kurtosis of a set of values indicates how flat or peaked the distribution is when compared to the normal distribution:
|
Value
|
Distribution Shape
|
|---|---|
|
kurt = 0
|
Normal distribution
|
|
kurt > 0
|
Relatively peaked distribution
|
|
kurt < 0
|
Relatively flat distribution
|
• skew(A, B, C, ...)—Returns the skewness of the elements A, B, C, .... The skewness is defined as follows:
The skewness of a set of values measures the asymmetry about the mean:
|
Value
|
Distribution Shape
|
|---|---|
|
skew = 0
|
The distribution is symmetric about its mean, as is the case for the normal distribution.
|
|
skew > 0
|
The distribution "tail" extends towards positive values.
|
|
skew < 0
|
The distribution "tail" extends towards negative values.
|
Arguments
• A, B, C, ... are scalars or arrays. There must be at least four elements in the arguments for kurt, and at least three for skew.
• M is an array created from the function arguments A, B, C, .... The standard deviation of M must not equal zero.