|
Estimation Commands
|
|
The Statistics package supports a variety of tools for manipulating likelihood functions, performing maximum likelihood estimation and deriving various properties of the likelihood function.
|
|
|
Examples
|
|
|
>
|
|
Calculate the likelihood, log likelihood and score function of a beta distribution
|
>
|
|
|
![PIECEWISE([0, A[1] < 0], [A[1]^(-1+a)*(1-A[1])^(-1+b)/Beta(a, b), A[1] < 1], [0, 1 <= A[1]])](/support/helpjp/helpview.aspx?si=9225/file02860/math110.png)
| (1) |
|
>
|
|
|
![ln(A[1]^(-1+a)*(1-A[1])^(-1+b)/Beta(a, b))](/support/helpjp/helpview.aspx?si=9225/file02860/math117.png)
| (2) |
|
>
|
|
|
-Psi(a)+Psi(a+b)], [ln(1-A[1])-Psi(b)+Psi(a+b)]])](/support/helpjp/helpview.aspx?si=9225/file02860/math124.png)
| (3) |
Attempt to calculate the maximum likelihood estimate of a binomial distribution.
|
>
|
|
|
![S := RTABLE(18446790986280507382, float[8], Vector[row], rectangular, Fortran_order, [], 1, 1 .. 1000)](/support/helpjp/helpview.aspx?si=9225/file02860/math133.png)
| (4) |
|
>
|
|
|
| (5) |
Calculate the same information about a normal distribution.
|
>
|
|
|
>
|
|
|
| (6) |
|
>
|
|
|
| (7) |
Compute the likelihood ratio statistic about a similar normal distribution.
|
>
|
|
|
![(1/25)*A[1]^2-(2/25)*A[1]*mu+(1/25)*mu^2](/support/helpjp/helpview.aspx?si=9225/file02860/math169.png)
| (8) |
|
|
Download Help Document
Was this information helpful?
