Sufficient statistics for the Bernoulli and Normal models

Sufficient statistics are far from unique; this returns the standard minimal choice.

Usage

Suffstat(x, family = "normal")

Arguments

x

The sample.

family

Either "bernoulli" or "normal".

Value

List with T1, T2, n, dim, mle_mu, mle_sigma2.

Details

Formula: Bernoulli T = sum_i x_i; Normal T = (xbar, s)

References

Wasserman (2004), All of Statistics, Section 9.13.2, Definition 9.32 and Examples 9.33 and 9.34.

Examples

V <- c(1, 2, 3, 4, 5, 6, 7, 8)
Suffstat(V)
#> $T1
#> [1] 4.5
#> 
#> $T2
#> [1] 2.44949
#> 
#> $n
#> [1] 8
#> 
#> $dim
#> [1] 2
#> 
#> $mle_mu
#> [1] 4.5
#> 
#> $mle_sigma2
#> [1] 5.25
#> 
#> $method
#> [1] "Normal sufficient statistic, Wasserman Ex 9.34"
#> 
#> attr(,"class")
#> [1] "morie_rich_result" "list"