Kurtosis (excess) of a data sample
MuPAD® notebooks are not recommended. Use MATLAB® live scripts instead.
MATLAB live scripts support most MuPAD functionality, though there are some differences. For more information, see Convert MuPAD Notebooks to MATLAB Live Scripts.
x1, x2, …) stats::kurtosis(
[x1, x2, …]) stats::kurtosis(
…, xn) returns
the kurtosis (the coefficient of excess)
where is the mean of the data xi.
The kurtosis measures whether a distribution is “flat” or “peaked”. For normally distributed data, the kurtosis is zero. If the distribution function of the data has a flatter top than the normal distribution, then the kurtosis is negative. The kurtosis is positive, if the distribution function has a high peak compared to the normal distribution.
We calculate the kurtosis of some values:
stats::kurtosis(0, 7, 7, 6, 6, 6, 5, 5, 4, 1)
Alternatively, data may be passed as a list:
stats::kurtosis([2, 2, 4, 6, 8, 10, 10])
We create a sample:
stats::sample([[a, 5, 8], [b, 3, 7], [c, d, 0]])
a 5 8 b 3 7 c d 0
The kurtosis of the second column is:
We create a sample consisting of one string column and one non-string column:
stats::sample([["1996", 1242], ["1997", 1353], ["1998", 1142]])
"1996" 1242 "1997" 1353 "1998" 1142
We compute the kurtosis of the second column. In this case this column does not have to be specified, since it is the only non-string column:
The statistical data: arithmetical expressions.
A sample of domain type
An integer representing a column index of the sample
FAIL is returned,
if the kurtosis does not exist.