summ_moment() computes a moment of distribution. It can be one of eight
kinds determined by the combination of central, standard, and absolute
boolean features. summ_skewness() and summ_kurtosis() are wrappers for
commonly used kinds of moments: third and forth order central standard ones.
Note that summ_kurtosis() by default computes excess kurtosis, i.e.
subtracts 3 from computed forth order moment.
summ_moment(f, order, central = FALSE, standard = FALSE, absolute = FALSE)
summ_skewness(f)
summ_kurtosis(f, excess = TRUE)| f | A pdqr-function representing distribution. |
|---|---|
| order | A single number representing order of a moment. Should be non-negative number (even fractional). |
| central | Whether to compute central moment (subtract mean of distribution). |
| standard | Whether to compute standard moment (divide by standard deviation of distribution). |
| absolute | Whether to compute absolute moment (take absolute value of
random variable created after possible effect of |
| excess | Whether to compute excess kurtosis (subtract 3 from third order
central standard moment). Default is |
A single number representing moment. If summ_sd(f) is zero and
standard is TRUE, then it is Inf; otherwise - finite number.
summ_center() for computing distribution's center, summ_spread()
for spread.
Other summary functions:
summ_center(),
summ_classmetric(),
summ_distance(),
summ_entropy(),
summ_hdr(),
summ_interval(),
summ_order(),
summ_prob_true(),
summ_pval(),
summ_quantile(),
summ_roc(),
summ_separation(),
summ_spread()
d_beta <- as_d(dbeta, shape1 = 2, shape2 = 1)
# The same as `summ_mean(d_beta)`
summ_moment(d_beta, order = 1)
#> [1] 0.6666667
# The same as `summ_var(d_beta)`
summ_moment(d_beta, order = 2, central = TRUE)
#> [1] 0.05555556
# Return the same number
summ_moment(d_beta, order = 3, central = TRUE, standard = TRUE)
#> [1] -0.5656854summ_skewness(d_beta)
#> [1] -0.5656854
# Return the same number representing non-excess kurtosis
summ_moment(d_beta, order = 4, central = TRUE, standard = TRUE)
#> [1] 2.4summ_kurtosis(d_beta, excess = FALSE)
#> [1] 2.4